University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Virginia Tech | Chapter 3: BOS Analysis of Ventilation Laboratory Fans
3.1 Description of Study
The following BOS study seeks to evaluate the potential of utilizing the BOS
technique for visualizing flow produced at a mine fan. The mine fan is the most
important part of a ventilation system as all airflow is produced by this component. As a
result, a considerable amount of pre-operational planning must be completed before the
fan can be installed. Once installed, surveys of mine fans are completed at regular
intervals to ensure proper operation. The ability to visualize airflows within the
immediate vicinity of the fan can provide insight into how efficiently flow is being routed
underground. Complex fan housing and ducting systems create complicated flow
patterns.
Tangible images of how the air actually flows in these types of systems can
indicate where to collect flow data and how to modify the design to minimize areas of
high resistance and stagnation. These images may also have application for the design of
fan housings such as evasés and diffusers. Laboratory testing must first be completed in
order to evaluate the feasibility of imaging airflow induced by comparable fans before
experiments can be completed in the field. The following experiment uses the BOS
technique to image airflow generated by a laboratory scale axial vane fan and a
laboratory scales axial flow fan. The BOS images are created using both an artificial
BOS background and a natural BOS background. The subsequent sections outline the
initial experimental efforts for visualizing flow produced by these two laboratory scale
fans.
35 |
Virginia Tech | 3.2.3 Equipment Setup
The two backgrounds were placed approximately 0.60 m (24 in) behind both fans.
The backgrounds were landscape oriented so that shortest side faced toward each fan‟s
inlet. The face of the background was oriented toward the camera lens. The Nikon
D5000 was mounted on the tripod and placed approximately 1.85 m (6.07 ft) in front of
the centerline of the two fans. The space heater was oriented so that its face points
toward each fan‟s inlet to allow the heated air to directly flow into the fans. The heater
was positioned approximately 0.6 m (24 in) from the face of both fans. These distances
were selected avoid obstructions present in the laboratory. The axial vane fan and the
axial flow fan are both permanently mounted in different areas of the lab. The axial vane
is installed on a ledge that is 1.5 m (57 in) in height. The axial flow fan is installed on the
floor in the center of the room. Figure 3.3 displays the layout of the laboratory as
viewed from the southeast corner of the room.
Figure 3.3. Photo of the ventilation laboratory as viewed from the southeast corner.
38 |
Virginia Tech | The previous figure clearly displays the heated air being drawn into the fan‟s inlet, which
matches the observation made in Figure 3.5. As previously stated, the heated air images
captured using the imitation rock background as well as the images captured without heat
using both backgrounds did not display flow. Figure 3.7 displays the post-processed
image generated by the axial flow fan using the heater against the rock background.
Figure 3.7. Schlieren image of heated airflow using the axial flow fan
against the imitation rock background.
As can be seen in Figure 3.7, no obvious flow is present in the image. The post-
processed image for the second fan and remaining background-heater combination are
not shown because these images also lacked flow.
3.4 Discussion
The images presented in Figures 3.5 through 3.7 demonstrate that the BOS
technique resulted in varying levels of success. The BOS technique was used to image
flow with two fans, an axial vane fan and a custom built axial flow fan. A space heater
was used in various trials to enhance visualization of the airflow. Two backgrounds
consisting of black and white stripes and an imitation rock pattern were used in
conjunction with the fans. Images utilizing the heater that were captured from both the
axial vane fan and axial flow fan are displayed in Figures 3.5 and 3.6 respectively.
These figures clearly display the heated flow moving into each fan intake. In contrast,
42 |
Virginia Tech | the trials executed with the heater using the rock background as well as without the heater
for both backgrounds were not able to fully render any flow. An example of this issue
can be seen in Figure 3.7.
This outcome shows that the BOS technique can clearly display airflow under the
correct conditions. As a result, the potential for large scale application of this technology
does exist. However, the lack of success using the imitation rock background in all trials
indicates that problems must be present in the ability of the background pattern to
enhance the schlieren effect. This problem can be present in one of three forms:
insufficient spatial frequency, insufficient contrast, or a combination of all four areas.
The spatial frequency and contrast of the background pattern allow distortions caused by
the schlieren effect to be made apparent when the flow image is correlated to the static
image. If the pattern is unable to reveal the differences between the images, the flow
cannot be visualized. The rock background may lack the spatial frequency and contrast
needed to visualize the airflow at this scale. This issue may also be a factor in the
inability of the striped background to visualize the isothermal airflow from the fans.
The large scale underground implementation of the BOS technique using this
form of the experimental design is not feasible. Despite these problems, the successful
production of some schlieren images warranted further investigation into the BOS
technique for use in underground mine ventilation. The experiments could not proceed in
the current study of isothermal airflow due to the insufficient imaging sensitivity of the
BOS setup. As a result, the research of this BOS investigation changed to encompass the
imaging of higher pressure flows through ventilation controls.
43 |
Virginia Tech | Chapter 4: Application of Qualitative BOS Analysis for Flow
through a Regulator
4.1 Description of Study
The following study seeks to evaluate the potential of the BOS technique for
imaging airflow through a regulator. Successful mine ventilation relies on ventilation
controls to operate at design specifications. If a single control malfunctions, the
effectiveness of the ventilation system can be drastically affected. Ventilation control
systems must be regularly maintained to ensure optimal effectiveness. Underground
environmental conditions, such as humidity, dust, ground movements, and water influx,
stress the integrity of ventilation controls. Visual inspections and regular maintenance
are currently the most effective means against this problem. However, even with regular
inspections, minor leaks can be missed due to the sheer volume of items that must be
examined. The best means of improving the performance of ventilation controls is to
increase the accuracy of inspections. This paper investigates a possible means of
attaining this objective through the BOS method. In addition to evaluating the imaging
ability of BOS, the experimental design will be concurrently validated. This validation
will be completed by comparing BOS images with traditional schlieren images. The
following sections contain a detailed description of the experimental setup and outlines
initial experimental efforts used for the investigation.
4.2 Experimental Design
4.2.1 BOS Equipment
The BOS technique in this experiment utilized a Nikon D700 digital camera, a
Nikkor AF-S 18 - 55 mm, and a BOS striped background. The camera imaged flow
induced by a 0.61 m (24 in) custom built axial flow fan. This fan is part of an exhausting
system and was attached to a 0.61 m (24 in) low turbulence wind tunnel. A sheet of
0.013 m (0.50 in) thickness plywood with a 0.051 m (2 in) diameter hole at its center was
44 |
Virginia Tech | 4.2.2 Single Mirror Schlieren Equipment
The traditional schlieren aspect of the following experiment utilized the single
mirror method. The experimented implemented a Nikon D700 camera with an AF-S VR
II Nikkor 300mm f/2.8G IF-ED super telephoto lens used in conjunction with a 0.15 m (6
in) optical grade concave spherical mirror with a focal length of 1.50 m (59 in). The
knife-edge obstruction consisted of a single, thin stainless steel plate. This plate was used
to partially block the beam from the point light source. The point light source was
custom built and consists of an LED capable of producing 180 lumens at 700 milliamps
(mA). This system will be used to image the same flow described in the previous section.
4.2.3 Experimental Setup
The setup for the BOS imaging experiment will be covered first. The plywood
regulator was affixed to the fan inlet so that the center of the hole aligned with the
centerline of the fan inlet. The schlieren background was placed approximately 0.46 m
(18 in) behind the centerline of the axial flow fan. The background was oriented so that
the longest side was parallel to the airflow. The stripes faced toward the camera‟s lens.
The Nikon D700, using a Nikkor AF-S 18 - 55 mm lens, was mounted on a tripod and
placed approximately 0.51 m (20 in) in front of the centerline of the fan. The space
heater was oriented so that its face (i.e. heat outlet) pointed toward the regulator hole.
This orientation allowed the heated air to flow directly into the regulator opening. The
heater was placed approximately 0.46 m (18 in) from the inlet of the fan. These distances
were selected based on setups used by previous researchers in conjunction with
constraints set by the architectural layout of the laboratory. The axial flow fan is installed
on the floor in the center of the room. Another wind tunnel is permanently mounted on
the floor directly adjacent to the axial flow fan. Figure 3.3 in Section 3.2.3 displays the
layout of the laboratory as viewed from the southeast corner of the room.
The research conducted in the BOS studies introduced in Section 2.5 suggests
that the highest quality BOS images are produced in one of two manners: capturing the
flow using a telephoto lens or imaging at close range with a short focal length lens (e.g.
46 |
Virginia Tech | that the face of the lens was located two mirror focal lengths, or 3 m (118 in), from the
face of the mirror. The LED was positioned next to the camera so that the front surface
of the bulb was flush with the front of the telephoto lens. Thus, the front of the light
source would also be located two focal lengths away from the concave mirror.
The beam was oriented so that it reflected off the center of the mirror and into the
center of the lens. The reflected beam thus projected an identical real image of the LED
onto the face of the lens. The knife-edge was then positioned in front of the lens in a
manner that intercepted the majority of the reflected light. The space heater was oriented
so that its face (i.e. heat outlet) pointed toward the regulator hole and fan inlet. This
orientation allowed the heated air to flow directly into the regulator opening. The heater
was positioned approximately 0.46 m (18 in) from the inlet of the fan. A diagram of the
single mirror setup can be seen in Figure 4.3.
Figure 4.3. Single mirror schlieren experimental setup.
48 |
Virginia Tech | The post-processed images from this experiment are displayed in Section 4.3. The single
mirror experimental procedures will now be discussed.
The camera was focused so that the mirror, regulator inlet, and heater were
framed in the picture. The system was calibrated using an active soldering iron. The iron
was placed approximately 0.051 m (2 in) in front of the mirror. The lens was then
adjusted so that the soldering iron was in focus. The knife-edge was adjusted until the
heat from the soldering iron could be clearly imaged. An example of the calibration
system can be seen in Figure 4.5.
Figure 4.5. Single mirror schlieren system being calibrated using
the heat distortion created by an active soldering iron.
Once the calibration was complete, the fan and heater were activated in various
combinations and captured with the camera. The resulting images are displayed in
following section.
4.3 Schlieren Imaging Results
The experiment delivered many different images with varying levels of success.
The BOS and traditional schlieren techniques used in this experiment were able to
visualize airflows that were enhanced by the space heater. The remaining trials in which
only the fan was active did not provide conclusive results. Out of the many BOS images,
two images will be discussed. These images accurately reflect the strengths and
50 |
Virginia Tech | 4.4 Discussion
The images presented in Figures 4.6 through 4.11 demonstrate that the BOS
technique resulted in varying levels of success in this experiment. BOS was used to
image airflow through a model plywood regulator induced by a custom built axial flow
fan. A space heater was used in various trials to enhance visualization of the airflow and
identify sensitivity problems. A single background consisting of alternating black and
white stripes served as the schlieren background for the Nikon D700 camera. A second
set of images using a single mirror schlieren system were captured to validate the BOS
technique designed for this experiment.
The BOS image of heated air is displayed in Figure 4.7. The heated
airflow image captured by the traditional system is shown in Figure 4.9. These figures
clearly display the heated flow moving into the regulator from the right. The single
mirror image is clearly comparable to the one created with the BOS method. This
similarity can be seen in the small vortices characteristic of turbulent flow that are
visualized in the traditional image. Therefore, the BOS images of the heated trials do
accurately represent the airflow through the regulator. Thus, this experimental technique
is validated. However, the result does not confirm the validity or feasibility of utilizing
BOS for analyzing unheated flows.
This conflict arises from the inability of BOS to fully render the unheated airflow.
An example of this issue can be seen in Figure 4.10. This result is confirmed by the
single mirror system image presented in Figure 4.11. The unsuccessful imaging of the
flow through the regulator is likely due to the sensitivity threshold of the laboratory
equipment. The pressure drop of 996 Pa (4.0 in H O) across the regulator did not create a
2
sufficient refractive gradient for both the BOS and single mirror schlieren systems.
Although auto compression from the reduction of inlet surface area does result in a
temperature change, the resulting refractive differential still needed to be enhanced by the
space heater. This outcome suggests that the airflow through the regulator can be imaged
even without the use of a heater as long as a large pressure differential exists (i.e.
surrounding air is colder than flowing air). The ability to image airflows was found to be
dependent on the consistency of atmospheric conditions.
54 |
Virginia Tech | Chapter 5: BOS Imaging of Methane Emissions
5.1 Methane Storage and Emission
Coal is produced from the accumulation of plant matter in very specialized
depositional environments. Over the course of geologic time, these layers of plant debris
can be exposed to certain magnitudes of pressure and temperature. If environmental
conditions are optimal for coalification, or the process by which coal is produced from
decomposing vegetation, the plant matter will be altered to form peat. Peat can then be
transformed into lignite, sub-bituminous, bituminous, semi-anthracite, anthracite, or
meta-anthracite coals depending on pressure, temperature, and time. The coalification
process is composed of a biochemical phase that is followed by a geochemical phase, or
alternatively, a metamorphic phase. During the geochemical or metamorphic stage, the
carbon content is increased while the hydrogen and oxygen contents are decreased. As a
result, methane, carbon dioxide, and water are produced as byproducts. Water is rapidly
lost while carbon dioxide and methane are retained in internal coal structures [3].
The stored methane exists as a free gas and as an adsorbed gas. The free gas is
contained in the pore spaces and fracture networks of the coal as freely moving
molecules. The remaining methane exists as adsorbed gas on the internal surface area of
the coal. This adsorbed methane becomes packed and stored as a monomolecular layer
on these surfaces. The great majority of methane is stored in this manner. Undisturbed
coal deposits naturally create a pressure equilibrium that prevents methane from
desorbing [2]. Gases can thus be indefinitely contained within in-situ coal as long as the
equilibrium exists.
The advancement of underground mine workings exposes coal to the atmosphere.
The bond between methane and coal is broken by the resulting pressure gradient. This
gradient causes methane to desorb from the coal [3]. Despite modern advancements in
quantitative methane monitoring technology, little is known about the qualitative aspects
of methane desorption. Questions about the manner in which methane desorbs from coal
are left unanswered due the highly dynamic mechanism of methane emission.
56 |
Virginia Tech | This complexity stems from the interactions between methane and coal‟s nano
and meso-structures. Coal can be classified as a solid that is composed of a multi-scale,
fractal-type, ramified internal surfaces. This type of internal structure contains a system
of pores, cracks, fissure, and voids that form intricate paths to the external boundaries of
the coal. The system as a whole can be referred to as a filtration volume in which
methane is pressurized as a result of natural compression and chemical interaction [65].
This random introduction of methane to coal causes the creation of completely
unique desorption potentials in all coal deposits. The pressure in undisturbed coal beds
can range from atmospheric at sea level to several megapascals (MPa) [65]. As a result,
desorption potential, or methane content, of different coal deposits can only be accurately
classified through physical data collection. Geophysical techniques do exist to survey
un-sampled coal deposits, but these methods are not as accurate. Although techniques
exist to quantify methane content, this property is different from the actual emission rate
of methane from exposed coal.
5.2 Discussion of Study
Emission rates have been modeled using numerical simulation techniques, but the
complexity of the methane migration mechanism still requires physical data to achieve
higher accuracy [66]. This complexity stems from the fact that methane emission rates
also encompass a kinetic aspect. Once a pressure differential appears, methane migrates
through coal by means of diffusion and filtration. Diffusion, or the movement of
molecules from high concentration areas to low concentration areas, occurs only over
short distances. Long-term mass transfer of methane is facilitated by diffusion. In
filtration, methane gas navigates through the internal system of pores, cracks, and voids
until it reaches the source of the pressure differential [65].
This process is driven by the severity of the pressure differential and the methane
replenishment ability of the coal. Furthermore, methane emission can be greatly affected
by the presence geologic anomalies, such as the degree of metamorphism, the depth of
bedding, and the occurrence of geologic structures (i.e. faults and intrusions) [66].
57 |
Virginia Tech | Methane desorption rates of Pittsburgh Coal in the size range of 180 to 200 mesh were
generated by the Bureau of Mines. The results of this study can be found in Figure 5.1.
Figure 5.1. Desorption curve for Pittsburgh Coal, 180 to 200 mesh [67].
The desorption cure shown in Figure 5.1 is unique to the Pittsburgh Coal sample
and size distribution used in the study. This curve will change with coal type, location,
and size [67]. Thus, the methane emission rates of coal deposits are highly variable and
difficult to predict. This study seeks to investigate the feasibility of using the BOS
technique for imaging methane flow.
5.3 Experimental Design
5.3.1 BOS Equipment
The BOS method for visualizing flow requires two pieces of equipment, a
professional quality digital camera coupled with an artificial schlieren background. For
this experiment, a Nikon D700 digital camera was used in conjunction with a BOS
striped patterned background. This pattern was selected for this experiment due to its
58 |
Virginia Tech | prior success in the BOS studies described in Chapters 3 and 4. The camera was used to
image pure methane flow from a laboratory grade methane gas cylinder to verify the BOS
technique for generating schlieren images in this experiment. The camera was mounted
to a tripod to ensure image stability and image consistency between the different pictures.
Once the procedure was validated, the BOS system was used to capture images of a setup
designed to simulate transparent inhomogeneous flow though a permeable material.
When rendered in aerosol at a barometric pressure of 101,325 Pa and a
temperature of 0°C, methane has a refractive index of approximately 1.000444 [68]. The
index will vary with different atmospheric conditions, which modifies the gas‟ density
distribution independently of flow. Under controlled laboratory conditions, the index is
not expected to change by more than ± 0.000156 according to the relationship displayed
in Equation 2.7 [69]. This phenomenon is discussed in Section 2.3.1. The BOS
background employed in the experiment will now be explained.
5.3.2 BOS Background
A striped BOS background was selected due to its success for imaging heated
airflows in the previously introduced BOS experiments. This background was composed
of a repeating pattern of 0.003 m (0.118 in) wide stripes whose colors alternate between
black and white. The striped pattern was drawn using a computer drafting program and
oriented so that they were parallel with the horizontal. The pattern was then plotted and
affixed to a foam poster board 0.81 m (32 in) by 1.02 m (40 in) in size. This pattern‟s
alternating light-dark contrasting color scheme created an ideal interface for BOS images.
An example of the striped background installed in the experimental setup can be seen in
Figure 5.2 on the following page.
59 |
Virginia Tech | The images of methane flow through permeable material were prepared by first
drilling a 0.0064 m (0.25 in) diameter hole in the center of a 0.070 m (2.75 in) diameter
sandstone core sample. The hole was drilled to a depth of approximately 0.070 m (2.75
in) into the 0.14 m (5.5 in) long core sample. A length of flexible tube was fully inserted
into the core sample through the drilled hole. The tube was then sealed in place with
ventilation grade silicon adhesive to prevent leakage from the immediate injection area.
When opened, the regulator injected methane into the center of the core sample. The
methane that exited the through the sandstone‟s porous openings was imaged by the BOS
system. Imaging methane flow through a coal sample would have been preferable, but
the friability of coal greatly increased the difficulty of the experimental setup. Sandstone
was chosen because it allowed for a reasonable flow of methane. The sample was also
easily drilled and sealed thus reducing the time needed for sample preparation. The core
sample placed in the schlieren image plane can be seen in Figure 5.4.
Airtight Sealant
Flexible Tube Outlet
Figure 5.4. Sandstone core sample placed within the BOS imaging system.
62 |
Virginia Tech | 5.3.4 Procedures
The outlet of the flexible tube was first placed in the BOS imaging area under the
fume hood. The camera was oriented so that the methane outlet and background were
framed in the photograph. Subsequently, the camera was focused so that both the BOS
background and the outlet of the flexible tube were clear. The static system was then
captured with the camera several times using various flash modes, aperture sizes, and
shutter speeds. In these photos, the methane cylinder was secured and the regulator was
closed. These initial pictures served as the reference images for BOS processing. All
other photographs were electronically correlated to the reference images.
Several photos were then taken with the gas regulator open and injecting methane
at an outlet pressure of 207,000 Pa (30 psi). Once the images were captured, the methane
cylinder was closed. The flexible tube was then disconnected and taken out of the fume
hood. The flexible tube and the sandstone core sample were then affixed to the methane
cylinder and placed in the image plane. As before, the static system was first captured.
These images were followed by a series of photos the regulator open at an outlet
pressures ranging from 138,000 Pa (20 psi) to 345,000 Pa (50 psi) to simulate flow
through a permeable material. These injection pressures were chosen in order to account
for the lower spectrum of pressures found in coal deposits so that the sensitivity threshold
of the BOS system could be evaluated [70]. After collecting the pictures from both
systems, the images were processed through GIMP v2.6, an open source image
manipulation program.
The flow image was first overlaid on the static image. The “difference” function
was then used to detect the variation between the static photo and the flow photo. This
function removes the colors of the selected photo from the background photo thereby
isolating the change in pixels from the static system to the dynamic system. The two
photos were then merged. The image contrast was enhanced using a Retinex function.
The Retinex algorithm enhances the visual rendering of an image in low lighting
conditions. Contrast curves were then adjusted to provide the optimum presentation of
the schlieren effect. A visual representation of this process can be seen in Figure 4.4.
63 |
Virginia Tech | As can be seen in the previous figure, the methane flow through the porous sandstone
core sample is not apparent to the naked eye. This result is identical to the image
displayed in Figure 5.5. Figure 5.8 shows the post-processed image of methane flow
through the porous sandstone generated in low light using an ISO of 1,000, an f-stop of
f/5, and the flash engaged.
Figure 5.8. Post-processed image of simulated methane desorption at an outlet pressure of 345,000 Pa
(50 psi).
The methane in Figure 5.8 is very apparent. The gas can be seen exiting the
sandstone core sample from the upper sides of the sandstone. The pattern exhibited by
the methane flow is similar to the pattern shown in Figure 5.6. However, the pattern in
Figure 5.6 is more characteristic with the vortices seen in turbulent flow. This difference
is consistent with the fact that methane flow through the sandstone is more tortuous and
is exiting from an increased surface area. As a result, the gas slowly escapes from the
injection area and out into the atmosphere. This behavior may more closely approximate
methane emissions from coal.
66 |
Virginia Tech | Chapter 6: Conclusions
This study evaluated the potential of background oriented schlieren (BOS) for
improving the data gathering process in underground mine ventilation surveys. This
investigation specifically concentrated on the qualitative data gathering potential of BOS
technology. Although quantitative techniques exist, this type of analysis was not pursued
due to the complex nature of the experimental design. The primary quantitative analysis
tool available for the evaluation of BOS images is particle image velocimetry (PIV). PIV
analysis is conducted by tracking the movement of turbulent structures from consecutive
BOS images. The tracking of particles is achieved through the use of custom developed
cross-correlation algorithms.
In order to apply PIV to BOS, the imaging system must be constructed using
exacting specifications. The imaging area must also follow such strict constraints so that
an accurate cross-correlation algorithm can be developed to represent the target flow.
The requirements of high precision and environmental control exclude PIV analysis as a
useful tool in underground mines. Additionally, BOS studies that have used PIV have
failed to produce useful quantitative results. This outcome is due to the relative
immaturity of PIV based BOS research. For these reasons, the use of PIV as an analysis
technique is eliminated for this study. A thorough discussion of PIV and its limitations
can be found in Section 2.5.4. The evaluation of BOS was completed in three main
phases: BOS imaging of two laboratory fans, BOS imaging of flow through a regulator,
and BOS imaging of methane emissions. These three experimental areas resulted in
varied levels of success and failure.
The first and second studies found that the flow induced by laboratory mine fans
and compressed flow through regulators respectively could not be imaged. The inability
of the BOS system to capture the flow was a result of either an insufficient refractive
index gradient or an inadequate refractive index contrast to the encompassing medium.
Thus, the sensitivity of this BOS system eliminates its potential to be a viable mine
airflow analysis tool. These unsuccessful imaging attempts prompted a major change in
focus for the BOS investigation to methane.
Methane flow was chosen due to its significance in underground coal mining and
its greater refractive potential as compared to air. The BOS system was able image
68 |
Virginia Tech | methane flow in the third experiment. The processed BOS images clearly displayed the
methane escaping from the sandstone core sample with an injection pressure of 345,000
Pa (50 psi). Ironically, the lower sensitivity threshold, although a limiting factor in the
earlier airflow studies, facilitated the imaging of methane. If BOS sensitivity were
sufficient to image all airflows, such as those created by the exhausting fume hood, the
appearance of these flows would have interfered with the visualization of methane. A
detailed discussion of this imaging study can be found in Chapter 5. The successful
production of BOS images confirms the feasibility of this technique for methane imaging.
However, the large scale underground implementation of the BOS technique in its current
form is not yet confirmed due to a lack of full scale underground imaging data. BOS
methane imaging research should be pursued further due to the inherent benefits that can
be gained.
At present, even with numerical models, the exact concept of how methane flows
from a mined surface of coal is unknown. Although the presence of methane and the
percent concentration in the air can be detected, the points at which methane flows from
the coal cannot be accurately ascertained. This aspect is important because of the basic
mechanics of methane release. Once coal is excavated, a large quantity of methane is
first released due to the exposure of coal to atmosphere and the resultant pressure
gradient between the reservoir and the atmosphere. However, the rate of methane
released fluctuates randomly due to the coalification and pressurization processes. Even
greater complexity is added considering that methane flow rates are affected by internal
pore structures, equilibrated pressure, coal particle size, and geologic structures [67].
Once the main methane plume is released, the flow continues but is instead fed by
filtration through coal channels. Methane release by desorption is highly variable thereby
making identification of release points improbable with quantitative techniques [65].
Qualitative schlieren analyses provide insight into these problems. BOS has the ability to
provide unique information about the release of methane into underground mines. This
technique can show how methane actually desorbs from coal as it is mined. This
technique can be used to identify any methane fissures on coal surfaces that are releasing
methane into the mine.
Once the characteristics of methane flow are qualitatively identified, the data can
be used to improve ventilation equipment designs. This improvement can include
69 |
Virginia Tech | optimizing the placement of methane sensors on continuous miners to increase the
effectiveness of early warning systems as well as enhance the placement of auxiliary
ventilation systems to better facilitate the dilution of methane. Additionally, BOS can be
used to demonstrate how the internal structural composition of coal and geologic
anomalies affect desorption mechanisms and emission rates of methane. The
aforementioned methane imaging experiments along with the two other experiments
conducted in this study have demonstrated the usefulness of BOS in underground mine
ventilation.
The first and second experiments showed that airflow with sufficient temperature
differentials from the surrounding atmosphere could be imaged effectively with BOS
techniques. Such differentials exist during certain seasons in the exhausting air from
underground mines. Successful imaging of exhausting air would most likely be achieved
during the peak summer or winter months where mine air is significantly colder or
warmer with respect to the atmosphere. The first experiment showed that a rock-type
pattern could not be used to image airflow in the laboratory. However, BOS imaging
using a rock background may still be successful if lighting conditions are modified to
enhance the contrast variability in the background. Furthermore, larger scale airflows
may be imaged with a rock background if the scale of the background is sufficiently
smaller than the flow pattern. A successful result could possibly be achieved when
imaging airflow from underground auxiliary ventilation systems.
The third experiment showed that clear images could be captured of methane
emissions even at low pressures. Such qualitative BOS information can be used in a
practical sense to optimize the procedures of ventilation surveys and design of ventilation
monitoring equipment. For example, images of methane flow in active mining areas can
be used to optimize the positioning of auxiliary ventilation equipment to dilute known
areas of high methane concentration. BOS images could also be used to re-evaluate the
placement of methane monitors on mining equipment to better facilitate the detection of
dangerous methane concentrations in active mining areas. For these reasons, further
investigation into the BOS technique for use in imaging underground airflows with
differential temperatures and methane emissions in underground coal mines is suggested
as an addendum to this study.
70 |
Virginia Tech | Chapter 7: Future Work
BOS is still relatively immature and has applicability in many fields. In mine
ventilation, two areas of improvement were chosen due to their helpfulness in advancing
the BOS technique introduced in this study: experimentation in an underground mine and
evaluation of background oriented optical tomographic techniques for mine ventilation.
In addition to the BOS experiments conducted in this study, the next logical progression
is to perform large scale field testing. This type of testing can be completed as a
concluding extension to the airflow studies in Chapters 4 and 5 as well as the methane
study introduced in the methane experiment discussed in Chapter 5. Underground field
testing can determine the viability of the BOS technique in coal mine ventilation.
Concurrent evaluations could also be made regarding the ability of BOS to image
methane against the natural backdrop of coal. The original experiment could not evaluate
the large scale viability of the BOS system in this manner due to the inability of the
laboratory setup to reproduce the exact characteristics of methane desorption from coal.
This limitation in the experimental design can thus be remedied by applying this
experiment‟s specific BOS technique to freshly mined coal. Successful imaging of this
scenario will serve as a more comprehensive evaluation. The second possible area of
study is optical tomography.
Tomography, in general, is an analysis technique that produces three-dimensional,
virtual reconstructions of the internal structure and composition of objects. This
reconstruction is created from the observation, recording, and examination of the passage
of energy waves or radiation through a target object. Tomography is a complementary
technology to already prevalent energy based observation devices. These devices include
radar, sonar, lidar, echographs, and seismographs. Tomographic systems use the
information gathered by the aforementioned systems to calculate physical parameters
with respect to the spatial information of the data. Tomography utilizes inverse theory to
image the interior of an object or mass [56]. This description encompasses a wide variety
of tomographic analysis methods. The one method of direct interest to BOS is optical
tomography.
71 |
Virginia Tech | Experimental Parameters Camera Settings
Shutter
Speed
Date Photo Lens AirflowHeater Lighting Scheme Flash Focal Plane ISO Aperture (s)
9/30/2010 DSC_1800 18 - 55 mm Lens Off Off No Ambient Light On Background 200 f/4.2 1/60
9/30/2010 DSC_1801 18 - 55 mm Lens On Off No Ambient Light On Background 200 f/4.2 1/60
9/30/2010 DSC_1802 18 - 55 mm Lens On On No Ambient Light On Background 200 f/4.2 1/60
9/30/2010 DSC_1803 18 - 55 mm Lens Off Off Overhead Fluorescent Off Background 200 f/4.2 1/15
9/30/2010 DSC_1804 18 - 55 mm Lens On Off Overhead Fluorescent Off Background 200 f/4.2 1/15
9/30/2010 DSC_1805 18 - 55 mm Lens On On Overhead Fluorescent Off Background 200 f/4.2 1/15
9/30/2010 DSC_1806 18 - 55 mm Lens Off Off No Ambient Light On Background 200 f/4.2 1/250
9/30/2010 DSC_1807 18 - 55 mm Lens On Off No Ambient Light On Background 200 f/4.2 1/250
9/30/2010 DSC_1809 18 - 55 mm Lens On On No Ambient Light On Background 200 f/4.2 1/250
9/30/2010 DSC_1820 18 - 55 mm Lens Off Off Overhead Fluorescent Off Background 6400 f/4.2 1/1250
9/30/2010 DSC_1821 18 - 55 mm Lens On Off Overhead Fluorescent Off Background 6400 f/4.2 1/1250
9/30/2010 DSC_1822 18 - 55 mm Lens On On Overhead Fluorescent Off Background 6400 f/4.2 1/1250
9/30/2010 DSC_1824 18 - 55 mm Lens Off Off Overhead Fluorescent Off Background 200 f/29 2
9/30/2010 DSC_1825 18 - 55 mm Lens On Off Overhead Fluorescent Off Background 200 f/29 2
9/30/2010 DSC_1826 18 - 55 mm Lens On On Overhead Fluorescent Off Background 200 f/29 2
9/30/2010 DSC_1827 18 - 55 mm Lens Off Off Overhead Fluorescent Off Background 200 f/4.2 1/250
9/30/2010 DSC_1828 18 - 55 mm Lens On Off Overhead Fluorescent Off Background 200 f/4.2 1/250
9/30/2010 DSC_1829 18 - 55 mm Lens On On Overhead Fluorescent Off Background 200 f/4.2 1/250
9/30/2010 DSC_1835 18 - 55 mm Lens Off Off Overhead Fluorescent Off Background 200 f/16 0.77
9/30/2010 DSC_1836 18 - 55 mm Lens On Off Overhead Fluorescent Off Background 200 f/16 0.77
9/30/2010 DSC_1837 18 - 55 mm Lens On On Overhead Fluorescent Off Background 200 f/16 0.77
9/30/2010 DSC_1838 18 - 55 mm Lens Off Off No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1840 18 - 55 mm Lens On Off No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1841 18 - 55 mm Lens On On No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1842 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1843 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1844 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1845 18 - 55 mm Lens Off Off No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1846 18 - 55 mm Lens On Off No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1847 18 - 55 mm Lens On On No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1848 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1849 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1850 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1851 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1852 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1853 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1854 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1855 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1856 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1857 18 - 55 mm Lens Off Off No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1858 18 - 55 mm Lens On Off No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1859 18 - 55 mm Lens On On No Ambient Light On Center of Regulator 200 f/4.2 1/60
9/30/2010 DSC_1860 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1861 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1862 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/15
9/30/2010 DSC_1863 18 - 55 mm Lens Off Off No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1864 18 - 55 mm Lens On Off No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1865 18 - 55 mm Lens On On No Ambient Light On Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1866 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1867 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1868 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 6400 f/4.2 1/1250
9/30/2010 DSC_1869 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1870 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1871 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/29 2
9/30/2010 DSC_1872 18 - 55 mm Lens Off Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1873 18 - 55 mm Lens On Off Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
9/30/2010 DSC_1874 18 - 55 mm Lens On On Overhead Fluorescent Off Center of Regulator 200 f/4.2 1/250
132 |
Virginia Tech | DEVELOPMENT OF A NOVEL FINE COAL
CLEANING AND DEWATERING TECHNOLOGY
Nikhil Gupta
ABSTRACT
The cleaning and dewatering of ultrafine (minus 44 micron) coal slurries is one of the
biggest challenges faced by the coal industry. Existing commercial technologies cannot produce
sellable products from these ultrafine streams; therefore, the industry is forced to discard this
potential energy resource to waste impoundments. This practice also has the potential to create
an environmental hazard associated with blackwater pollution. To address these issues,
researchers at Virginia Tech have worked over the past decade to develop a novel separation
process that simultaneously removes both mineral matter and surface moisture from fine coal
particles. The first stage of the process uses immiscible non-polar liquids, such as straight chain
hydrocarbons, to selectively agglomerate fine coal particles in an aqueous medium. The
agglomerates are then passed through the second stage of processing where mild agitation is used
to disperse and fully engulf hydrophobic coal particles into the non-polar liquid and to
simultaneously reject any residual water and associated hydrophillic minerals entrapped in the
agglomerates. The non-polar liquid, which has a low heat of evaporation, is then recovered by
evaporation/condensation and recycled back through the process. The research work described in
this document focused on the engineering development of this innovative process using batch
laboratory and continuous bench-scale systems. The resulting data was used to design a proof-of-
concept (POC) pilot-scale plant that was constructed and successfully demonstrated using a
variety of fine coal feedstocks. |
Virginia Tech | ACKNOWLEDGEMENTS
I would like to express my deepest gratitude to my adviser and my mentor, Dr. Gerald H.
Luttrell and Dr. Roe-Hoan Yoon, for their continuous guidance, criticisms and valuable insight
throughout this research. The valuable comments and continuous support of the committee
members, Dr. Stanley C. Suboleski and Dr. Gregory T. Adel, are also gratefully acknowledged.
The financial support from the U.S. Department of Energy and Mineral Refining
Company, LLC (Evan Energy, LLC) is greatly appreciated.
I am heartily thankful to Mr. Robert (Bob) Bratton for his endless support and
motivation, which helped me to complete my research successfully. I acknowledge great support
from Mr. Jim Waddell for his tireless work in construction of the pilot plant. I am thankful for
the Department of Environmental Health and Safety in guiding us to operate the pilot plant in
intrinsically safe manner. I extend my appreciation to the excellent staff of the Department of
Mining and Minerals Engineering, Ms. Katheryn Dew, Ms. Carol Trutt, and Ms. Gwen Davis for
their administrative support in all manners.
I am thankful to the highly supportive staff of Arch Coal, especially the Lone Mountain
facility, for providing samples for pilot-scale testing whenever needed. Assistance from Consol
Energy and Alpha Natural Resources for providing samples is also appreciated.
I wish to thanks my parents, Mrs. Manju Gupta and Mr. Akhilesh K. Gupta, for their
unconditional love and support. I would not be able to achieve this milestone without their
prayers and blessings.
Finally, I express my deepest appreciation to my beautiful wife, Alba R. Cordova, for her
love, patience, and strong support. She and her family have been very understanding and
encouraging for me in completing my research successfully.
iii |
Virginia Tech | CHAPTER 1 – General Introduction
PREAMBLE
Coal has been fulfilling a high proportion of human energy needs from centuries. Though
historically used as a domestic fuel, coal is now globally used by industries, especially in the
generation of electricity. Coal is the largest and historically one of the cheapest fuels used for
electricity generation in the United States and accounted for 37% of electric power generation in
2012. Since 2000, about 90% of all the coal consumed in United States has been used for electric
power generation (EIA, 2013). The U.S. Energy Information Administration expects total
consumption will increase by 7.1% from its current value in the next couple of years (refer
Figure 1.1) due to higher electricity demand and increasing natural gas prices.
Coal mining grew aggressively during the Industrial Revolution, which started in the
1880’s. Since then the mining practices have improved from men tunneling, digging and
manually extracting the coal on carts to large surface and underground longwall mines. Mining
at high production volume requires modern large machines such as draglines, trucks, conveyors,
self-advancing longwall supports and high-capacity shearers. The large-scale extraction of coal
through the 20th century in the traditional coalfields of the Eastern United States has resulted in a
diminution of the mined coal quality, while mechanization has resulted in reduction in average
particle size. These factors, in addition to the 1970’s “energy crisis” and strict environmental
standards, have forced the United States coal industry to find ways to produce marketable coal in
a more efficient manner. Coal preparation, a process that improves the quality of coal and
recovers coal particles from coal-rock run-of-mine material, has become more important as the
mined product has worsened.
Coal preparation processes that remove unwanted impurities and increase the product
heating content significantly improve coal quality. This methodology results in increased power
plant capacity and reduced plant maintenance cost. Raw coal that emerges from a mine contains
“bone” (clay/shale rock), pyrite, mercury and other types of impurities; these vary in amount
depending on the particular coal seam and mining method used. Physical processes can remove
many of the included non-combustible mineral impurities, whereas chemical processes can
remove impurities, such as organically bound sulfur, that are part of the
1 |
Virginia Tech | EIA (2013), “Short-Term Energy Outlook”, U.S Energy Information Administration, Report – June 2013
Figure 1.1 U.S. electricity generation (thousands MWH/day) by fuel (EIA, 2013).
complex chemical composition of coal and are impossible to remove by physical methods. In
short, coal processing increases the heating value, lowers the transport cost per unit of heat,
reduces emissions, and improves the marketability of the run of mined coal. There are currently
286 coal processing plants in the United States that clean approximately 67% (660 million short
tons per year) of the total coal consumed in the United States (Coal Age, 2010).
Moisture is considered to be a contaminant in the final clean coal products. Excess
surface moisture also reduces the heating value of coal, and can lead to severe handling and
freezing problems. Additionally, it also increases the transportation cost of coal. Relative to
energy loss, moisture and ash content are roughly equivalent (Luttrell, 2010), which is
approximately 150 BTU loss for each additional 1% by weight.
2 |
Virginia Tech | LITERATURE REVIEW
The following literature review is designed to provide a brief overview of the current
accepted practices for fine coal cleaning and dewatering circuits in coal preparation plants.
Moreover, the literature review provides detailed background on the oil agglomeration process,
which is a crucial aspect of the reported research. In addition, an overview of the previous
scientific studies, which are the foundation of proposed technology, conducted at Virginia Tech
is presented. In the final section, the research objectives and goals are discussed in detail.
1.1 Coal Preparation Practices – An overview
The earliest type of coal preparation employed were “hand pickers” to remove non-coal
materials from coal. With the advent of mechanization, more sophisticated techniques were
necessary to clean large amounts of smaller, more impure coal particles. As a result, coal
preparation plant technology has evolved in close harmony with changing mining technology
and practices.
Modern plants are designed in accordance with specific operational factors like raw coal
characteristics, market specifications and demands, environmental requirements, applicable
processing methods, and economics. Each of these factors dictates the role of different unit
operations, sizing, cleaning, dewatering and drying in the final design of the coal preparation
plant. A typical coal processing plant flowsheet can be represented by a series of sequential unit
operations for coal particle sizing, cleaning and dewatering. This sequence of operations is called
a circuit. Coal processing operations must be designed in multi-stage circuits for several different
size fractions, since each coal preparation method has a limited range of applicability in terms of
particle size. In the United States, processing plants typically include as many as four separate
processing circuits for treating the coarse coal (above 10 mm), intermediate (10 x 1 mm), small
(1 x 0.15 mm) and fine (below 0.15 mm) feed material. Figure 1.2 illustrates a typical flowsheet
for a modern coal preparation facility.
The coarse coal cleaning processes typically involve Jigs or Dense Medium Vessels. For
material in range of 10 x 1 mm, dense medium processes are used to efficiently clean run-of-
mine coals, while screens and centrifugal dryers are used to efficiently dewater the clean coal
products. Both coarse and intermediate coal particle separations are based on differences in the
3 |
Virginia Tech | relative densities (RD) of coal (1.3 RD) and associated impurities (2.0 RD). Particles in the size
class between 1 and 0.15 mm are typically cleaned using water-based density concentrators
including spirals, water-only cyclones, crossflow/teeter-bed separators or multi-stage
combinations of these units. The only commercially viable process for treating particles finer
than 0.15 mm is froth flotation. Particles smaller than 1 mm, which are more difficult to dewater
due to a higher specific surface area, typically require the use of energy intensive devices such as
screenbowl centrifuges or filters to remove the unwanted surface moisture. The problems
associated with fine coal cleaning processes are complex. To address the issue and increase the
plant productivity in this circuit, an emerging practice is to deslime the flotation feed using a
classifying cyclone. The cyclone separates ultrafines (below 0.044 mm), which cannot be
processed economically with the existing technologies and are thus discarded. Therefore, it is
now paramount of importance to recover coal from these streams by developing a suitable
method.
Figure 1.2 Generic modern coal washing plant flowsheet
4 |
Virginia Tech | 1.2 Fine Coal Cleaning
Effective cleaning of fine coal (that is, both removal of impurities and moisture
reduction) is mostly dependent on the economics, the capability and performance efficiency of
the processing equipment, and the extent to which separation of the feed coal can be optimized.
Currently, froth flotation is the only commercially practiced method for cleaning ultrafine coal in
the United States. In previous years, oil agglomeration was another extensively studied method
for fine coal cleaning, but could not grow and lost the significance due to several reasons. Due to
its high cleaning and to some extent dewatering capabilities and its importance to this research, it
is reviewed in detail in a separate section of this chapter.
1.2.1 Conventional Froth Flotation Process
Froth flotation is currently the preferred method for cleaning fine coal particles of size
below 150 microns (minus 100 mesh). It is based on the differential wettability of particles; i.e.,
this surface-based process distinguishes between hydrophobic coal and hydrophilic mineral
impurities (clay, pyrite etc.). In flotation, air bubbles are dispersed in water in which fine coal
and mineral matter are suspended. Hydrophobic coal particles are selectively collected by a
rising stream of air bubbles and form a froth phase on the surface of the aqueous phase, leaving
the hydrophilic mineral matter behind. Higher-rank coal particles are usually hydrophobic and,
therefore, can be attracted to air bubbles that are also hydrophobic through a mechanism known
as hydrophobic interaction. Along with surface chemistry, particle and bubble size are two of the
most important variables. Flotation works best for fine particles about 0.1-0.25 mm in diameter.
Larger particles (greater than 0.25 mm) have a high probability of bubble-particle detachment,
whereas smaller ones (less than 0.1 mm) have a low probability of bubble-particle collision.
While particle size determines which particles are most likely to float, bubble size
controls the amount of particles that are able to float. The total surface area of the bubbles
determines the carrying capacity of the froth. Therefore, smaller the bubble size, the greater the
bubble-particle interaction. Several advanced flotation technologies have been successfully
commercialized. Nonetheless, their primary focus is to create smaller bubbles inside the
flotation cell. In previous years, Virginia Tech has successfully commercialized micro-bubble
column technology (MicrocellTM), which showed high carrying capacity and energy recovery
(Yoon et al., 1992) because of micron-size bubbles.
5 |
Virginia Tech | In a modern coal flotation circuit, the feed stream (minus 100 mesh) is first classified
using a 6-inch classifying cyclone to remove ultrafines of size below 44 microns (minus 325
mesh). When the clean coal product reports to the froth phase, it is substantially free of mineral
matter but contains a large amount of water. Finer particles have greater surface and greater
capacity to adsorb water. Wet fine coal is difficult to handle, increases shipping costs and lowers
combustion efficiencies. Therefore, the clean coal product is dewatered using various devices
such as cyclones, thickeners, filters, centrifuges, and/or thermal dryers.
1.2.2 Problems with Fine Coal Processing
There are two reasons for the high costs of processing fine coals (0.15 x 0.044 mm): one
is the low efficiency of cleaning, and the other is associated with the high cost of dewatering.
The first problem has been resolved to a large extent by the advent of advanced coal cleaning
technologies, such as the conventional and column flotation process, and advanced flotation
methods such as MicrocellTM column flotation, StackCell® flotation, etc. These water-based
processes are capable of recovering the fine coal from finely dispersed ash-and SO -forming
2
minerals; however, it is difficult to remove the free water adhering to the surfaces of fine coal
particles. The finer the particle, the larger the surface area and, hence, the more difficult it
becomes to dewater. Typically, 100 x 325 mesh flotation concentrate contains 30-40% moisture
after a mechanical dewatering process such as vacuum filtration, causing not only a loss of
heating value, but also problems with handling and transportation. Some consider that cleaning
fine coal replaces one type of inert substance (e.g., ash-forming minerals) by another (e.g.,
water), offering no financial incentives for coal companies to clean fine coals (Yoon and Luttrell,
1995).
In general, the cost of dewatering increases with decreasing particle size (as illustrated in
Figure 1.3) and can become prohibitive with ultrafine particles, e.g., particles finer than 0.044
mm (minus 325 mesh). In such cases, coal producers are forced to discard those because of
unacceptably high moisture content and processing cost. The top size of the material discarded may
vary from 0.15 to 0.044 mm (i.e. 100 to 325 mesh) depending on the value of the coal and demands
imposed by the sales contract (NETL, 2009). Ultrafine coal is one of the primary components of
the fine waste found in waste impoundments. The loss of the minus 44 micron material is
especially tragic because coal particles that small are liberated extremely well. Large amounts of
6 |
Virginia Tech | Figure 1.3 Effect of particle size on dewatering cost ($/ton), (redeveloped, first published by
Hucko in 1981)
fine coal have been discarded to numerous imp oundments worldwide, creating environmental
concerns. A study conducted by National Research Council under congressional mandate reports
some 70 – 90 million tons of fine coal is being discarded to coal slurry impoundments annually
by the United States coal industry (Orr, 2002). The industry estimates that so far approximately 2
billion tons of fine coal have been discarded in abandoned ponds, and 500-800 million tons are
in active tailing ponds (Orr, 2002). This activity represents a loss of a valuable energy resource,
loss of profit for coal producers, and the creation of a potential environmental liability.
1.3 Fine Coal Dewatering Methods
The solid-solid separation processes employed by modern coal preparation plants require
large amounts of process water. After cleaning, the unwanted water must be removed from the
surfaces of the particles. Small and fine coal (less than 1 mm) particles represent as little as 10%
of the total run of mine coal and often contain one-third or more of the total moisture in the final
coal product (Osborne, 1988). Existing fine coal dewatering processes, such as filtration,
centrifuges, and thermal drying are expensive, inefficient, and consume a lot of energy (Osborne,
1988).
7 |
Virginia Tech | Figure 1.4 Dewatering methods with respect to size fraction of coal
Coarse coal particles larger than 5 mm are dewatered using screens. Shaking and
vibrating screens are commonly used for this purpose. Moreover, sieve bends are generally used
for preliminary dewatering of coal prior to vibrating screens and centrifuges. Particles of 5 x 0.5
mm size range are typically sent to basket type centrifuges for dewatering.
In this research document, three conventional methods of dewatering used for small size
fraction are reviewed: screenbowl centrifugation, vacuum filtration, and thermal drying (refer
Figure 1.4). Screen bowl centrifuges are widely used in coal industry to dewater the 1 mm x 0
size range of clean coal coming from froth flotation and spirals. Screenbowls are able to handle
some ultrafine sizes and, therefore, are reviewed here, although they are usually reserved for
coarser feeds than those studied in this research. If high coal recovery is desirable, then the fine
coal (0.5 mm x 0) can be dewatered using vacuum filters. Vacuum filtration is the most common
method for dewatering ultrafines. Although thermal drying produces the driest product, it is
currently the least used of the three methods due to high cost and problems in obtaining
environmental permits. Less than 10% of the existing United States coal plants still utilize
thermal dryers for moisture control (Bratton, 2013), largely because of abovementioned issues.
Latest developments and emerging technologies for dewatering and drying solids, such as
8 |
Virginia Tech | hyperbaric centrifugation technology, Nano-Drying method, etc., will also be discussed in the
section.
1.3.1 Centrifuges
Centrifugal dewatering is a solid-liquid separation technique in which solid particles are
separated from a liquid by means of a combination of sedimentation and filtration mechanisms
driven by centrifugal force. These devices spin either horizontally or vertically. The rotation
generates centrifugal force, which separates water from fine coal, much like the spin-dry cycle of
a laundry washing machine (Osborne, 1988).
Although gravitational sedimentation and centrifugation employ the same basic principle,
i.e. differential density separation, the latter is a much faster process because of the centrifugal
‘g’ force applied to the particles. Most of the centrifugal dewatering devices used in coal industry
operate at 50-3000 times the gravitational force. High g-forces cause solids to settle quickly into
a compact cake and force water out through the pores (Osborne, 1988). Two types of centrifuge
are commonly used in industry: solid-bowl and screen-bowl.
Bowl type centrifuges were first used in the coal industry in the mid-1960s with the
introduction of the solid-bowl centrifuge. These centrifuges contain two rotating elements: the
conveyor and the bowl. The bowl consists of a long cylindrical shaped region and a shorter cone
shaped region. The conveyor, with one or more helical flights that follow the contour of the
bowl, transports the material by rotating at a slightly slower or faster speed than the bowl
(Osborne, 1988). The unit can have either a concurrent or a countercurrent feed arrangement. In
the concurrent feed system, the pulp enters the centrifuge at the larger cylindrical section of the
bowl, and the cake moves in the same direction as the effluent towards the conical end.
Concurrent (solid-bowl type) operates at slower speeds as compare to countercurrent type,
therefore they are found to be attractive in removal of coal tailings where acceptable product
moistures are 35-45% in range.
Screenbowl centrifuges, as exhibited in Figure 1.5, are countercurrent machines and
consist of a horizontal tube with a screw inside to move the material. The pulp enters the
centrifuge near where the conical section starts and the cake moves in opposite to the effluent
flow. The machine is equipped with an additional cylindrical screen that assists further water
drainage. The first section of the horizontal tube is solid and removes the bulk of the water.
9 |
Virginia Tech | Decanter Machine Inc. (2013), www.decantermachines.com
Figure 1.5 Screenbowl centrifuge section diagram (Decanter Machine Inc., 2013)
The screen is made of tungsten carbide b ars that have wedge profiles to prevent near size
solid particles from getting stuck between the bars. As the feed comes into the horizontal tube
section, it quickly forms a cake while the majority of the liquid and about half of the minus 325
mesh material flow over the adjustable weirs in the back of the machine (Keles, 2010). Solids
settled under the acceleration force are carried up the slope of the cone by the helical conveyor,
as in solid-bowl centrifuges. However, unlike solid-bowl centrifuges, thickened cake of solids
pass over the cylindrical screen section where the remaining excess water is filtered through the
cake and discarded (Osborne, 1988).
These centrifuges are high capacity, long life machines that can provide low moistures.
The final moisture is directly related to the amount of minus 325 mesh feed material. For
example, if a feed contains 30% minus 325 mesh, the product’s moisture will be around 18%
(Osborne, 1988). It should also be noted that some of this ultrafine material is discarded with the
main effluent. Typically this effluent is not recycled, and any material in it is lost to the tailings.
Final product moisture is also dependent on the centrifugal force. A higher operating speed will
lead to lower moisture and a finer cut; however, screen-bowl centrifuges are generally not
operated above 500g due to excessive wear (Osborne, 1988). Due to the strong dependence of
product moisture on feed size distribution and limited centrifugal force, screen bowl centrifuges
are generally used for dewatering fine material coming off of spirals, i.e. 1 x 0.15 mm.
10 |
Virginia Tech | 1.3.2 Vacuum Disk Filtration
Filtration is used to separate liquids from solids by passing the solid-liquid mixture
through a permeable medium. The medium accumulates most of the solid particles while
allowing the liquid to pass. In coal preparation applications, most are disc-filters utilizing
vacuum and positive air pressure as the collection and dewatering mechanism.
Vacuum filtration is a highly effective method for dewatering fine coal containing a
large proportion of minus 325 mesh (minus 44 micron) solids. These filters are basically porous
cloth or fine-fabric screens to which a vacuum is applied. The vacuum draws water and solids to
the screen surface, which traps the solids but allows the water to pass through.
The most common type in the United States is disc-filters (illustrated in Figure 1.6),
which consist of vertical discs with fan shaped sectors covered in fine cloth. The hollow discs are
under vacuum and submerged about half way in slurry. As the discs rotate, they pick up solids
from the slurry. The cake dries as it is carried through the air, and then the dried cake is blown
off by positive pressure before the segment is again dipped into the slurry (Osborne, 1988). Fine
solids are trapped in the cake against the filter cloth, with recovery exceeding 97%. Moisture is
typically in the 25-35% range, and reagents may be needed to reach the lower moistures.
Flocculants are usually added to reduce screen blinding, reduce ultrafine losses, and aid in cake
release, while cationic coagulants are occasionally used to increase the filtration rate.
These filters are popular because of their small footprint, high capacity, and low cost;
however, they produce higher moisture levels and require more maintenance compared to some
other filters. Other continuous vacuum filters include rotary drums and horizontal belt filters.
Filtration may also be done by applying positive pressure instead of a vacuum; however, these
filters are more expensive and are used rarely in the coal industry for dewatering clean coal
products.
11 |
Virginia Tech | NFM (2013), National Filter Media, www.nfm-filter.com
Figure 1.6 Schematics of typical vacuum disc filters (NFM, 2013)
1.3.3 Hyperbaric Centrifuge System
One of the latest centrifugal bowl type separators is the Hyperbaric Centrifuge
(commercially known as CentribaricTM Centrifuge), which was developed at Virginia Tech for
ultrafine particle dewatering. The technology combines centrifugation and pressure filtration
techniques within one process to substantially reduce product moisture. Keles et al. (2010)
performed moisture-recovery analysis on a prototype hyperbaric filter centrifuge unit
manufactured by Decanter Machines. The moisture values were reported in the range of 13 to
20% with solid recoveries in range of 83-96%. It was demonstrated through economic analysis
that utilization of hyperbaric centrifuges in a coal plant would likely produce an attractive
economic gain compared to utilizing only screen bowl centrifuges (Keles, 2010). The first
commercial hyperbaric centrifuge unit (Figure 1.7), also manufactured by Decanter Machines,
Inc., was tested by Walter Energy in 2009 by replacing the standard screenbowl centrifuge.
The most economic benefit observed on the commercial scale was the reduction in the
amount of ultrafine solids reporting to the centrifuge through the effluent in the plant. The
percent of solids reported was between 0.5 – 1%, with an ash value ranging from 30 to 50% as
compared to 4 – 6% solids with ash value 14% from previously installed screenbowl centrifuge
12 |
Virginia Tech | A Timed
Slurry
0.1 mm Screen Housing Rotary
Opening Panel Valve
Bars Cake Port Drain Ports Pneumatic
Cylinder
Air
Pressure Cake
Rotation
Chamber
Filter Feed
Cake Inlet Sealing Scraper
Edge Seal
Belt
Drive
Motor
Wedge Bars
Cross Section A-A
Cake Drain
Discharge A Discharge
Keles, S. (2010), PhD Dissertation, Mining& Minerals Engineering, Virginia Tech
Figure 1.7 Simplified schematics of hyperbaric filter centrifuge (Keles, 2010)
main effluent. This improved the plant productivity to 20 - 25 tons/hour (Franklin at el., 2012).
T he only drawback with this technology is that it is highly energy intensive, and thus very costly
to implement for a low price commodity such as coal.
1.4 Fine Coal Drying Methods
1.4.1 Thermal Drying
Thermal drying is not common in the United States as it is the most expensive unit
operation in coal preparation (Osborne, 1988). Additionally, it is now extremely difficult to get
permits to install new units in the preparation plants (Bratton, 2013). They are generally used on
ultrafine coals whose large surface areas lead to high moisture contents. Thermal dryers are the
only units that can consistently provide low moisture (<10%) with ultrafine feed, although they
diminish the coking properties of coal. The coal begins to volatilize at temperatures greater than
90°C, while these units typically operate in range of 150-430°C. Thermal drying is justifiable
only if the low moisture is worth the cost to reduce the possibility of freezing, to reduce heat loss
during combustion, and to prepare the coal for coke making (Osborne, 1988).
13 |
Virginia Tech | Jumah, R, and Majumdar, A, (2006), “Dryer Emission Control System”, Handbook of Industrial Drying
Figure 1.8 Simplified flowsheet of fluidized bed direct heat exchanger coal dryer
Industrial coal dryers usually employ convection in direct heat exchange type dryers. In
these type dryers wet coal is continuously brought into contact with hot gases in order to
evaporate surface moisture (Osborne, 1988). The most common types of convective thermal
dryers are: fluidized bed, flash, and multi-louvered (Jumah and Majumdar, 2006). In the
fluidized bed dryer (outlined in Figure 1.8), the coal is suspended and dried above a perforated
plate by rising hot gases. In the flash dryer, coal is fed into a stream of hot gases for
instantaneous drying. The dried coal and wet gases are both drawn up a drying column and into a
cyclone for separation. In the multi-louvered dryer, hot gases are passed through a falling curtain
of coal, which carried by flights of specially designed conveyor.
1.4.2 Parsepco Drying Technology
Mohanty et al. (2012) reviewed several emerging fine coal drying technologies, of which
Parsepco Drying Technology (PDT) is one of them. PDT employed medium–wave infrared
radiation (MIR) in combination with a steel-belt dryer and a pin mixer (as described in Figure
1.9). This infrared drying system transfers thermal energy to the feed material (typically 25-30%
moisture) to be dried in the form of electromagnetic waves, producing dry product that is
constantly below 10% moisture level. Buisman (2010) indicated medium-wave infrared radiation
is more effective than short-wave or long-wave infrared in moisture reduction. A pin mixer is
used for ultrafine clean coal below 75 microns to prepare the
14 |
Virginia Tech | Buisman, R. (2010), “Coal Fine Beneficiation Using Liquid Coal Fuel”, IQPC Presentation
Figure 1.9 Schematic of Parsepco Drying Technology (Buisman, 2010)
feed for the dryer. It is reported that moisture l evels in the range of 9-13% were achieved by
drying product obtained from a plate-and-frame filter press (Buisman, 2010).
1.4.3 Drycol Process
The Drycol Process was first developed by DBAGlobal Australia. The process utilizes
controlled application of microwave radiation for drying the fine coal fraction. Water molecules
attached to the coal surface absorb energy from the radiation by dielectric heating—that is, by
rotating rapidly as they attempt to align themselves with the microwave induced alternating
electric field. The schematic for the process is shown in Figure 1.10. The molecular movement
creates heat as the rotating molecules strike other molecules and put them in motion (Graham,
2007). The applied microwave energy passes through the carbon and acts directly on both free
and inherent water.
Microwave drying is well known for its advantages, such as volumetric heating and faster
drying rates. A commercial unit of capacity 15 tons per hour plant was tested and was able to dry
low-rank coal from 28% to 12% moisture content (Graham, 2007).
15 |
Virginia Tech | Graham, J. (2007), “Microwaves for Coal Quality Improvement: The Drycol Project”, DBAGlobal Au.
Figure 1.10 Simplified Drycol Process flowsheet (Graham, 2007)
1.4.4 Nano Drying Technology
The Nano-Drying Technology (NDT™) system uses molecular sieves to wick water
away from wet fine coal particles and does not require crushing or additional finer sizing of the
wet coal to dry it. These molecular sieves are a form of nano-technology based particles, which
are typically used for extracting moisture from airborne, aerosol and liquid environments. There
are also known techniques for combining molecular sieves with solids, but no previous
techniques included regeneration of the molecular seives. Molecular sieves contain pores of a
precise and uniform size. These pores are large enough to draw in and absorb water molecules,
but small enough to prevent any of the fine coal particles from entering the sieves. Some
molecular sieves can absorb up to 42% of their weight in water (Bland et al., 2011). Molecular
sieves are used in the drying process because they are re-usable after the absorbed water is
removed from the sieves by heating them in the microwave at 300°C (Bratton, 2013). The water
in the molecular sieve turns into vapor at this temperature and is released into the atmosphere.
Bratton et al. (2012) conducted both bench scale and pilot scale parametric testing and
statistical analysis on this technology. The study reported product moisture contents in the range
of 5% to 10% for both minus 0.6 mm and minus 0.15 mm fine coal material from the feed
carrying moisture in the range of 22% to 28%. A simplified schematics of the process is
exhibited in Figure 1.11.
16 |
Virginia Tech | Bratton, R., Ali, Z., Luttrell, G., Bland, R., and McDaniel, B. (2012), “Nano Drying Technology: A New
Approach for Fine Coal Dewatering”, 29th Annual ICPC, Lexington,, KY.
Figure 1.11 Simplified flowsheet for pilot scale NDTTM process (Bratton et al., 2012)
1.5 Oil Agglomeration
Another fine coal cleaning process that has been investigated in the past is selective oil-
agglomeration. Several studies conducted on oil agglomeration process for fine coal achieved
better combustible recoveries compared to conventional flotation process. The process is not
preferred in the United States due to high costs associated with oil consumption, among other
factors. The process is discussed in detail due to its importance in the novel technology proposed
in this research.
1.5.1 Brief History
Oil Agglomeration, which is also referred to as selective agglomeration or spherical
agglomeration, was first performed on coal in the early 1920’s (Mehrotra et al., 1983); however,
it was not until the 1970’s energy crisis that the United States invested significant amounts of
time and money into the potential uses of oil agglomeration. Though most of the testing during
the 1980’s focused on the cleaning ability of oil agglomeration, dewatering, and oil recovery
were also explored to some extent. Since its introduction in 1921, the use of oil agglomeration
was mostly investigated in laboratories and pilot plants. Several variations of oil agglomeration
processes were developed over the years including the Trent Process (1921), the Convertol
Process (1952), NRCC (1961), the Shell Pelletizing Separator Process (1968), the Olifloc
Process (1977), the CFRI Process (1976), and the BHP process (1977). Several pilot plants were
even constructed to test the feasibility of the method in continuous larger scale production
17 |
Virginia Tech | (Mehrotra et al., 1983). Most of abovementioned processes used light diesel oil as an
agglomerants that could not be easily recovered, and thus increased the cost of the final product.
To address the issue, Smith and Keller (1981) employed fluoro-chloro derivatives of
methane and ethane, which have low boiling points (40-159oF), so that the agglomerants could
be readily recovered and recycled by gentle heating. However, these reagents are known to have
undesirable effects on the atmospheric ozone layer, therefore could not be used on a large scale
production. Keller (1984) disclosed a method (called the Otisca T- process) of oil agglomeration
that utilizes short-chain hydrocarbons, such as, 2-methyl butane, pentane, and heptane as
agglomerants. Agglomeration using low chain recoverable non-polar liquids as an agglomerant
typically provides moisture only up to 40% by weight (Keller, 1985). These reagents also have
relatively low boiling points, allowing them to be recycled. Being able to recycle an agglomerant
is a significant step towards commercialization of the selective agglomeration process (Keller,
1984).
Another method to substantially reduce the amount of oil consumption was proposed by
Capes (1989). In this low-oil agglomeration process, smaller agglomerates (<1 mm) are formed
at low dosages of oil (0.5-5%) and are separated from mineral matter by flotation process rather
than by screening. Similarly, Wheelock and Meiyu (2003) developed another method of
selectively agglomerating coal using microscopic gas bubbles to limit the oil consumption to 0.3-
3% by weight of coal.
Figure 1.12 Interfacial forces on solid particle at oil-water interface
18 |
Virginia Tech | Chiang and Klinzing (1986) disclosed a method (called the LICADO process) for
cleaning fine coal of mineral matter by selective transportation of particles across the
water/liquid carbon dioxide interface. Additionally, the liquid CO could be recycled. A report
2
shows that clean coal products obtained using this liquid carbon dioxide process contained 5-
15% moisture after filtration (Cooper et al., 1990).
1.5.2 Fundamentals
The selective oil agglomeration process is a solid/solid separation method. The treatment
of coal fines in the aqueous suspension consists of separating the carbonaceous fraction from the
ash-forming mineral matter. The separation technique in oil agglomeration involves the principle
of preferential wetting of hydrophobic carbonaceous particles by hydrophobic liquid/oil. In the
presence of an adequate amount of oil and sufficient mechanical agitation, the oil coated coal
particles collide with each other and form into agglomerates. The agglomerate formation is due
to the interfacial tension of the oil and the capillary attraction of the oil bridges between the
particles. The mechanism of particle absorption by the bridging oil is explained in Figure 1.12.
The position of the solid particle at the interface is governed by the relative values of interfacial
tensions. From the balance of forces, only the following three conditions are possible, (a) if θ <
90°, the particle will tend to be drawn into the aqueous phase; (b) if θ = 90°, the particle will
remain at the interface; (c) if θ > 90°, the particle will be drawn into the oil phase (Osborne,
1988). Having the angle θ, defined as the contact angle, greater than 90° is the prime condition
for a successful oil agglomeration process.
Despite the fact that oil agglomeration has been studied extensively, the microscopic
interactions are still not well understood. Coal is not-homogenous and consists of a patchwork of
hydrophilic and hydrophobic sites (Keller and Burry, 1987); therefore, several conflicting
theories exist on which liquid, oil, or water, acts as the bridging mechanism to form the
agglomerates.
The first popular theory is that oil acts as a liquid bridge between coal particles (Keller
and Burry, 1987). In the oil agglomeration process, oil is added to an aqueous suspension while
being agitated. Under conditions of high-shear agitation, the oil breaks up into small droplets that
collide with coal particles, spread on the surface of the particles, and form pendular bridges
between them to produce agglomerates. The oil envelops the coal and bridges over the
19 |
Virginia Tech | hydrophilic sites. Though small droplets of water may remain bound to the hydrophilic sites, oil
displaces the water from the hydrophobic sites and remains the dominant liquid in the
agglomerates. As two oil-coated particles collide during mixing, the oil and its capillary
attraction causes the particles to stick together and eventually form agglomerates.
A second opposing theory suggests that water actually acts as the bridging liquid. Many
oils simply spread on hydrophobic coal surfaces, whereas water sticking to the hydrophilic sites
forms water droplets with contact angles greater than 90° (Good and Islam, 1991). When two of
these droplets meet, they form a bridge and the surface tension of the water pulls the coal
particles together. The more the particles are pulled apart, the more the surface tension increases
and forces these particles back together. In contrast to the first theory, hydrophobic liquids will
break apart into two droplets when the bridge is stretched (Good and Islam, 1991). Oil simply
coats the particles and provides an environment for the water bridges. There is little discussion
on whether these theories are mutually exclusive or may both contribute to agglomerate
formation.
1.5.3 Parameters
In the selective oil agglomeration process, the interaction that occurs between the
hydrophobic particles and hydrophobic liquid, which also affects the kinetics of the process, is
mainly controlled by three factors: (a) the free energy at the three-phase interface (the interface
between water, coal, and the hydrophobic liquid), (b) the dosage of hydrophobic liquid, and (c)
the mixing intensity (Capes and Darcovich, 1984).
Agglomeration should proceed when there is sufficient driving force, that is, the free
energy is negative (Keller and Burry, 1987). This is highly dependent on the nature and choice of
agglomerant used and the quality of coal in the process, and is thus governed by the interfacial
surface tension between water, coal, and the oil. The higher the rank of coal the higher is the
hydrophobicity of the coal surface. The thermodynamic model (proposed by Jacques et al., 1979)
based on the oil bridging mechanism showed the relationship between the change in free energy
with interfacial tensions with the following equation:
( ) ( )
[1.1]
20 |
Virginia Tech | Where F and F are the free energies of State I (when coal particle and oil droplet are fully
I II
dispersed in water phase) and State II (when coal particle is completely engulfed with oil in
water phase, agglomerated state) respectively, ϒ ϒ and ϒ are the interfacial tensions (refer
o-s w-s o-w
Figure 1.12) and n is the ratio of the diameter of particle in State II to the diameter of particle in
State I.
Previous studies (Keller and Burry, 1987; Skarvelakis et al., 1995) showed that coal
cleaning decreases with the increase in density and viscosity of oil. This is due to the decrease in
the interfacial tension between water and oil interface. The researchers found that chlorinated
hydrocarbons and short-chain alkanes were more effective than saturated aliphatic hydrocarbons,
which, at the same time, were better than aromatic hydrocarbons. The detailed analysis of the
feasibility of the oil agglomeration process for all rank coals (anthracite, bituminous, sub-
bituminous, and lignite) with these various types of agglomerants can be found in the literature
(Keller and Burry, 1987; Skarvelakis et al., 1995).
Moisture held by agglomerates of fine coal is mostly due to the amount of water trapped
between the void space in agglomerate and the moisture held on the particle surface. The amount
of oil required cannot be determined without consideration of the mixing intensity. Intensive
mixing such as that produced by high shear devices often generates rapid agglomerates through
efficient oil dispersion and good particle contact. The early Trent Process employed low intensity
agitation with high oil dosages and longer retention time to achieve agglomeration: on the
contrary, the Shell process used high mixing speed and low retention time for agglomeration.
Later processes, such as the Convertol and the Olifloc, used the combination of very high shear
mixing, low oil dosages (2-7% by weight of coal) and very short retention time (Capes and
Darcovich, 1984).
The next most important parameter for a successful agglomeration process is oil dosage.
The amounts of oil used in the process are typically in the range of 5 to 30% by weight of feed
coal (Tsai, 1982). At low dosages, agglomerates have larger void spaces between the particles,
forming agglomerates that are filled-up with water. Here, fine mineral matter, e.g., clay, is
dispersed, which makes it difficult to obtain low-moisture and low-ash products. Researchers
found indeed that the moisture content was in excess of 50% by weight when the amount of oil
used was less than 5%. By increasing the oil dosage to 15%, the agglomerates were more
compact and discreet in nature. They grew in size and were easy to drain the clay and mineral
21 |
Virginia Tech | matters with the suspending water phase. The void space in the interior of agglomerates was
partially filled with oil, which resulted in a lower moisture and cleaner product. A dosage higher
than 20% led to the formation of relatively hard and spherical agglomerates (Capes and
Darcovich, 1984).
Keller and Burry (1987) increased the dosage of oil to 55-56% by volume to fill up the
void spaces thoroughly, which practically eliminated the entrapment problem and produced
super-clean coal containing less than 1-2% ash. However, the moisture content remained high.
Keller (1985) also claimed that the typical agglomerates’ moisture content was 40% by weight
using fluorocarbons as an agglomerant. Depending on the type of coal tested, approximately 7-
30% of the moisture was due to the water adhering onto the surface of coal, while the rest was
due to the massive water globules trapped in the agglomerates (Keller and Burry, 1990).
1.5.4 Kinetics
Agglomerates growth kinetics depends on the aforementioned process variables and thus
affects the recovery efficiency of the coal as well as its moisture content. Many studies were
previously conducted (Rao and Vanangamudi, 1984; Skarvelakis et al., 1995) to determine the
kinetics and mechanism for the batch-scale oil agglomeration process and to predict the size
distribution of the agglomerates. The researchers showed the agglomerate growth rate follows
second order kinetics and can be represented with the following equation:
[1.2]
Where, d is the size which allows 50% agglomerates to pass, d is the equilibrium size that
50 50∞
can be obtained after a prolonged period of the agglomerating process, t is the agglomeration
period and K is the second order rate constant. The knowledge of the two constants, d and K ,
2 50∞ 2
for a given set of conditions, allows the growth of agglomerates as a function of agglomeration
time and mean diameter of the coal particle to be predicted.
Thermodynamically, the kinetics of the agglomeration process is faster than the flotation
method. This is because in flotation when an air-bubble contacts a particle its curvature changes.
This creates an excess pressure (P) on the wetting film existed between the bubble-particle. The
excess pressure due to curvature change (P ) is known as Laplace or capillary pressure. This
cur
pressure causes film thinning only up to a critical thickness. At critical thickness, electrical
22 |
Virginia Tech | double layer and van der Waals forces interacts with each other and give rise to a disjoining
pressure (Π). A pressure balance along the direction normal to a film shows that the excess
pressure becomes equal to the capillary pressure minus disjoining pressure (P = P – Π). In
cur
flotation conditions, both electrical double layer and van der Waals forces are repulsive
(positive), causing the excess pressure to decrease and hence the film thinning process de-
escalated (Sulman et al., 1905).
On the contrary, in oil agglomeration process the van der Waals forces are attractive
(negative) while electrical double layer forces are negligible in the presence of non-polar liquid
(oil). The negative disjoining forces increases the excess pressure on wetting film, thus escalates
the film-thinning process beyond the critical thickness. Furthermore, Pan and Yoon (2010)
identified that the higher the hydrophobicity of the particle (such as high-rank coals) the higher is
the negative disjoining pressure. For the reasons, oil agglomeration process has faster kinetics
and thermodynamically more favorable than flotation.
1.6 Otisca T-Process
The concept of Otisca T-Process is the key step in the proposed novel fine coal cleaning
and dewatering technology and thus is separately discussed in detail. The process was disclosed
by Keller (1984) and developed by Otisca Industries, Ltd. of Syracuse, NY. The process first
employed heavy chloro-flouro carbon (CFC) derivatives (1.25 – 1.7 SG), which have low boiling
point and can be recovered by heating at low temperatures.
The Otisca process involved three steps: (a) particle size reduction of run-of-mine coal to
15 micron x 0 in presence of chloro-flouro carbon, where the organic liquid forms a thin surface
film on the newly exposed particle surface, (b) agglomeration of the carbonaceous material from
coal-mineral matter-liquid system and separation of the agglomerates by gravity in a static bath
using chloro-flouro carbon as a medium, and (c) organic liquid recovery from both the clean coal
product and reject (Keller, 1982; Keller and Rainis, 1980). A simplified flowsheet of the process
is shown in Figure 1.13.
Keller (1982) showed that the process, when treating the grinded ultrafine size fraction
(15 micron x 0), was able to achieve almost 100% carbon recovery (ash value 0.3%) from run-
of-mine feed treated on a 200 pound/hour plant. The water content in the final product was
reported as low as 8% by weight. Later, the company constructed and successfully operated a 15
23 |
Virginia Tech | ton/hour feed capacity pilot-scale facility with Island Creek Coal Co. in Bayard, West Virginia in
late-1970s. The data collected from the plant showed energy recovery as high as 90% (Keller,
1982). The separation efficiency indicator, called the ecart probable, values obtained from 8-hour
pilot-tests for different size fractions are outlined in Table 1.1. The results clearly indicated that
as the size fraction decreases the ecart probable value increases (i.e., process efficiency
decreases). However, the separation was much better in comparison to the other processing
methods for fine (100 x 325 mesh) particle size fraction (Keller, 1982). In addition, the pilot-
scale testing showed only 0.1% loss of organic liquid.
Table 1.1 Ecart probable values for Otisca Process at different size fraction feed
(Keller, 1982)
Feed Ecart Organic
Size Fraction Probable Efficiency%
3/8 x ¼ inches 0.008 100
¼ inches x 28 mesh 0.015 99
28 x 100 mesh 0.175 98
100 x 325 mesh 0.260 96
3/8 inches x 325 mesh 0.023 98
Keller, D, and Rainis, A., (1980), “Processes of Recovering Coal”, US Patent No. 4186887
Figure 1.13 Simplified Otisca T-Process flowsheet (Keller and Rainis, 1980)
24 |
Virginia Tech | Following the initial success, the company constructed the first full-scale plant for
American Electric Power (AEP) of rated capacity 125 tons/hour in early-1980s. The project
received a setback when it was determined that the organic liquid recovery process was not
economical and the plant was losing as much as 5% of chloro-flouro carbon, which escalated the
clean coal product cost. Furthermore, scientists discovered that these organic liquids were a
significant source of depleting the ozone layer in the earth atmosphere (Seaman, 1992). The
company lost contract with AEP.
Keller (1984) switched the chloro-flouro carbons to short chain alkanes (such as pentane)
in the process, and was able to produce a final product of similar carbon recovery. The Otisca
Industries received couple of contracts from Florida Power & Light and General Electric
respectively (Seaman, 1992). The process was only able to survive few more years, because in
late-1980s, the energy crisis was over. The cost of the final Otisca product could not compete
with the falling prices of oil. Later, the process utilized fine coal-water slurries from the
preparation plants to produce high carbon recovery product, but the moisture (as-received)
reported was high, on an average 40% by weight (Keller, 1985). Eventually, the Otisca-T process
lost its significance and abandoned in early-1990s.
The Otisca process using short-chain hydrocarbons is a source of inspiration in
developing the novel, innovative technology proposed in the research. Since surface forces get
stronger for micron size particles, the researchers at Virginia Tech have developed an additional
proprietary step to treat the 40% moisture agglomerated product, which can provide very low
moisture as well as high combustible recovery at a very low energy input. The innovative
method simultaneously cleans and dewaters ultrafine coal slurries by exploiting hydrophobic-
hydrophilic surface properties of the particles with a hydrophobic liquid.
1.7 Foundation of Novel Proposed Technology
Consistent higher energy recoveries achieved in the oil agglomeration in comparison to
flotation process and the successful demonstration of the Otisca T- Process using recoverable
straight chain hydrocarbons provided motivation to the researchers at Virginia Tech to develop
an innovative method for dewatering fine coal using the recyclable non-polar liquids. The
dewatering is achieved by allowing the liquids to displace surface moisture. The agglomeration
process has been expanded through research at Virginia Tech by developing an additional
25 |
Virginia Tech | processing step, i.e., the phase inversion step (from water-oil-water to oil-water-oil), which is
capable of ‘drying’ (dewater) the fine clean coal at room temperature.
Yoon and Luttrell (1995) first disclosed the concept of dewatering-by-displacement
(DBD) or hydrophobic-displacement, which is the foundation of the novel proposed technology.
The researchers claimed that the method is capable of achieving the same level of moisture
reduction as thermal drying at substantially lower energy costs, but did not mention the removal
of mineral matter from coal. The beauty of the DBD method is its thermodynamic spontaneity in
behavior compared to the thermal drying process, which is forced drying. The only energy
requirement in the DBD method is the recovery of hydrophobic liquid, which can be achieved by
gentle heating depending on the nature of the liquid. Figure 1.14 illustrates thermodynamic
comparison between the DBD and thermal drying methods. In the latter, large amount of heat
exceeding the latent heat of evaporation is required to remove all the water molecules which are
deposited in multilayers. On the other hand, in the DBD method the only energy required is just
to displace the water molecules in the bottom-most monolayer. The additional advantage of the
novel concept is that the volatile matter is retained, thus it does not change the coal properties.
Further, the explosion hazard is reduced, since high temperature heating is not involved.
Figure 1.14 Thermodynamic comparison between thermal drying and the DBD process
26 |
Virginia Tech | 1.7.1 Concept of Hydrophobic Displacement
The scientific evidences of hydrophobic forces were first measured and reported in the
literature by Israelachvili and Pashley (1984). The research showed that these are attractive
forces that generate between non-polar molecules in the presence of water. In case of high-rank
coal, the coal particles in water are hydrophobic in nature and therefore, have high affinity
towards the hydrophobic liquids, such as hydrocarbons. Due to the attractive hydrophobic forces
between the two, the liquid quickly engulfs the coal particle and displaces the surface moisture.
On the contrary, the clay is hydrophilic in nature, and therefore does not interact with the
hydrophobic liquid. Since the novel concept of dewatering was driven by the hydrophobic
interactions, it is also referred as hydrophobic displacement.
Dewatering-by-displacement (DbD) is a method of cleaning fine coal from its mineral
matter and simultaneously dewatering the clean coal product by displacing the water adhering to
the coal surface with a hydrophobic liquid. The displacement is achieved by using the phase
inversion process (Yoon et al. 2011). Use of such a liquid allows coal particles to be engulfed (or
transported) into the hydrophobic liquid phase, leaving hydrophilic mineral matter in the aqueous
phase.
In order for the displacement to occur spontaneously, the thermodynamic analysis
comparing a beginning state of coal (1) in water (3) and an end state of coal in a hydrophobic
liquid (2) is shown in Figure 1.15. Application of Young’s equation (by Thomas Young in 1805)
yielded the following criteria for thermodynamic spontaneous dewatering. The change in Gibbs
free energy (G) of displacement with respect to contact area (A) must be less than the difference
between surface free energies at the coal/oil interface (ϒ ) and at the coal/water interface (ϒ ).
12 13
ΔG /ΔA = ϒ - ϒ < 0 [1.3]
displacement 12 13
Furthermore, from thermodynamic equilibrium condition, shown in Figure 1.15, the following
relationship can be established:
ϒ - ϒ = ϒ Cosθ [1.4]
12 13 23
Therefore, from Equations 1.3 and 1.4, the condition for displacement can be re-established as:
ΔG /ΔA = ϒ Cosθ < 0 [1.5]
displacement 23
27 |
Virginia Tech | √ [1.6]
√ [1.7]
√ [1.8]
where superscript d refers to dispersion component of surface tensions. The equations work very
well for non-polar liquids and solid surfaces (Sohn et al., 1997).
Figure 1.16 shows the relationship between calculated contact angle (θ) and carbon
number. The contact angles of hydrocarbon liquid in water increased as the carbon number
decreased. Liquefied butane (C4) had the greatest contact angle at 110°; therefore, displacement
of water by liquid butane is thermodynamically most favorable. Pentane had the next highest
three-phase contact angle, which is 106° (refer Figure 1.16).
With the availability of three phase equilibrium contact angle, it is possible to measure
the change in free energy per unit area between the two states as illustrated in the Figure 1.15 by
the relationship established from the Dupre equation (by Lewis Dupre in 1869).
ΔG = ΔA (ϒ - ϒ )+ ΔA ϒ Cosθ [1.9]
dis 13 12 23
Therefore,
ΔG /ΔA = ϒ - ϒ + ϒ Cosθ [1.10]
dis 13 12 23
Also, the work per unit area required for displacement can be determined from the
following thermodynamic calculations. The work of adhesion, the amount of work energy per
unit area required to pull apart two phases/species (suppose A and B) in contact with each other
in presence of third phase, is given by the following equation (by Harkins in 1928).
W = ϒ + ϒ – ϒ [1.11]
adhesion (A-B) A B AB
Similarly, work of cohesion, the amount of work energy per unit area required to pull apart
single species in terms of its interfacial tension, can be written as:
W = 2ϒ , as ϒ = 0 [1.12]
cohesion (A) A AA
From the above relationships, the amount of energy required to pull one liquid in presence of
other liquid on the coal surface can be calculated, given that the equilibrium contact angle and
surface tensions are available. The lower the energy (i.e. more negative free energy), the more
thermodynamically favorable the process will be.
29 |
Virginia Tech | In the thermodynamic states, illustrated in Figure 1.15, the work required per unit area to
remove water drops (3) from the coal surface (1) in any medium (2) will be,
W = W + W –W – W [1.13]
321 13 22 12 23
Therefore, using Equations 1.4, 1.11, to 1.13, work per unit area for displacing a water droplet
from the coal surface can simply be described as:
W = ϒ (1 + Cosθ) [1.14]
321 23
Equations 1.9 – 1.14 were later used in the thermodynamic energy calculations for the process,
which was investigated in detail during the early phase of the reported research and therefore,
will be described in Chapter 3.
1.7.2 Previous Research at Virginia Tech
Studies in dewatering by displacement were initiated at Virginia Tech in 1995 and
included thermodynamic analysis and batch-scale testing with liquid butane. Sohn et al. (1997)
conducted batch-scale testing on a mid-volatile bituminous coal with liquefied butane
(pressurized 25-35 psig at room temperature) due to its large three-phase contact angle and ease
of recovery (boiling point, 30.2°F). When clean coal slurry was gently agitated with large
amounts of butane in a pressurized vessel, the resulting dry coal powder gathered on top of the
water phase. The concentrate (approximately 2 grams) was removed from the top of the powder,
and the initial weight for the moisture was taken after the sample sat at room temperature for 90
minutes. Testing indicated this was the approximate amount of time needed for butane to
evaporate. The best moisture, i.e. 1%, was reported with a butane-to-coal mass ratio of 2.0, a
solid content of 5%, and a settling time of 10 minutes. Initial testing showed that the butane
recovery would be high due to ease of evaporation and the minor loss of butane in the water.
Yoon et al. (2011) reported that significant amounts of the process water could be
entrained into the organic phase in the form of large water globules stabilized by hydrophobic
coal particles. It is well known that particle such as coal in oil, with three phase contact angles
larger than 90o, stabilize water drops in the oil phase. This stabilization of water leads to the
formation of water-in-oil emulsion (Binks, 2002). In general, the hydrophobic liquid containing
dry coal particles and entrained water in the form of water-in-oil emulsion is phase-separated
naturally from the aqueous phase containing hydrophilic mineral matter. This hydrophobic liquid
30 |
Virginia Tech | can be transferred to a size-size separator, such as screen, classifier, and/or cyclone, to remove
the globules of water from the dry coal particles (Yoon et. al., 2011).
Smith (2008) conducted extensive laboratory-scale bench test investigation to examine
several hydrophobic displacement (separation and dewatering) methods of oil agglomerated
products with liquid n-pentane. The methods included: hand shaking, screening, air
classification, centrifugation, filtration and displacement. The research was conducted to identify
conditions for stable agglomerates and procedures to evaluate pentane loss/consumption from
evaporation curves. The major parameter studied was the pentane-to-coal mass ratio, varying
from 0.11 to 1.99. It was reported that spherical agglomerates (formed when the pentane-coal
ratio was between 0.21-0.34) responded most efficiently for dewatering purposes by
hydrophobic displacement. Very high moisture was reported in all the methods when the pentane
dosage increased to the ratio higher than 1. This may be due to the formation of thick curd-esque
stable water-in-oil emulsions with coal, as identified by Capes and Darkovich (1984).
The hand-shaking method was performed continuously for five minutes to achieve
consistent results (Smith, 2008). The investigator reported that the lowest moisture observed was
16% by weight, with the formation of loosely-bound floc-like agglomerates. In addition, the
combustible recovery was higher than 90% irrespective of the oil dosage. In the screening
method, floc-like agglomerates were passed through a coarse sieve. This method was an
innovative concept because the dry solids associated with the agglomerates were passed through
the screen while the water globules stabilized with coal particles retained on the top of the
screen. Shaking of the sieve caused the small water droplets to coalesce and roll over the sieve.
The coating of coal prevented the coalesced water from wetting the screen. The lowest moisture
reported was 6.5% but the recovery decreased drastically to only 30% with this method (Smith,
2008).
An air-classification method was implemented briefly as described in the corresponding
study (Smith, 2008). The method utilized spherical agglomerates, and air was used to remove the
top agglomerate layer floating on the top of aqueous phase. The method was not developed
further due to the unpredictability of throwing water into the agglomerates, because of high air
pressure. The research indicated that the method was not draining any water trapped in the voids
of the agglomerates structure, thus retaining the moisture in the final product.
31 |
Virginia Tech | According to the Smith (2008), the centrifugation method appeared to be the best method
both in terms of lower moisture and high recovery product. The lower moisture was attributed to
an increase in centrifugal g-force with higher rotation speed (Capes and Darkovich, 1984). The
product moisture observed in the bench scale experiments was as low as 7.5%, and recovery was
always greater than 90% with a centrifugation spin time of 1 minute at 3280 RPM and a pentane-
coal ratio 0.32. The higher dosage resulted in agglomerates clumping and sticking together when
being fed to the centrifuge.
When the vacuum filtration method was employed during the investigation (Smith, 2008)
for dewatering pre-cleaned agglomerates, the filtration of these agglomerates resulted in
moistures in the range of 20-32%. Lower moisture was observed on samples containing less
ultrafines material and more coarse solids (Smith, 2008). The research also reported that the
higher moisture values were caused by the small water droplets in agglomerate voids that were
retained in the filter cake.
In the displacement method, Smith (2008) filtered the homogeneous coal slurry with
large amounts of pentane without using any oil agglomeration process step. The experimental
study provided the product moisture content in the range of 22-28% by weight, when pentane
was poured on the top of coal slurry phase being filtered. In theory, the liquid pentane should
displace the last droplets of water as it filters through the cake; however this was not observed
during the bench-scale experiments (Smith, 2008). The lowest moisture reported was 19.7%
when the vacuum pressure increased to 24mmHg with a drying time of 1 minute.
In 2010, a low-temperature drying process was developed at Virginia Tech to reduce the
moisture of coal agglomerates. The technology was applicable to coal agglomerates and filtered
flotation concentrate with less than approximately 22% moisture (Freeland, 2010). Three devices
were developed to explore the process: a static breaker, an air jet conveyor, and a centrifugal fan.
In each device, the coal agglomerates or cake were subjected to a high, mechanical shearing
force. Compared to the other two methods, the centrifugal fan consistently produced a low-
moisture product (less than 2%) without plugging (Freeland, 2010).
The newly developed low-temperature drying technique required a high amount of
airflow to dry the particles. The relative humidity and temperature of the ambient air have a large
impact on the water carrying capacity of the air. It was discovered that the process worked best
by heating the air to at least 48.9 ºC (120ºF) (Freeland, 2010). Unfortunately, heating the air
32 |
Virginia Tech | added an additional cost to the process. Based on the economic model developed to calculate the
cost of an industrial scale low-temperature dryer unit, it was discovered that the thermal dryer
requires $0.18/ton less energy-cost than the low-temperature drying technology (Freeland 2010).
Smith (2012) continued the study on the aforementioned innovative screening method to
achieve low moisture product (< 10 %) and high coal recovery by implementing multiple stages
of screening. Apparently, multi-stage screening had no effect on the reduction of moisture and in
improving combustible recovery. Next, Smith (2012) attempted to utilize a Teflon-coated mesh
to prevent wetting of the sieve with coal-coated water. The study showed that the Teflon mesh
works better for a brief period of time; however, as the shaking continues, the water droplets
coalesce together and with coal particles and eventually result in a thick sticky coal-mass on the
screen.
Although the innovative screening method produced single digit moisture, the screen size
area requirement and low recovery made it almost impossible to develop the process in a
practical industrial setting. In addition, blinding of the screen and the risk of wetting the sieve
could not be easily controlled in the plant environment (Smith, 2012).
Smith (2012) later employed two critical modifications that assisted in the development
of technology proposed in this research. First, a cylindrical column reactor was introduced after
mixing (formation of emulsions) step. The column was initially filled with clean water and then
with pure pentane to create a distinct two-phase system. The liquid pentane floated on the clean
water due to its lower density. Second, an ultrasonic source of energy was implemented on a
batch-scale, which was used for the dispersion of coal water-in-oil emulsions in bulk pentane
liquid column. The coal-water in oil emulsions was first formed by intensive mixing in a kitchen
blender with a high dosage of pentane. The overflow of the kitchen blender from a custom made
port, which was mainly emulsions, was pumped into the separate column reactor. An ultrasonic
probe, operated at a high frequency (20 KHz), was inserted below the oil-water interface in
aqueous phase from the bottom of the liquid-filled column reactor. As the emulsions were
pumped into the column, they started settling at the interface. The ultrasonic energy dispersed
these emulsions at the interface leaving hydrophobic coal particles in the pentane column and
releasing trapped globules with associated clay into the water phase. The dispersed coal in
pentane was collected from the overflow port of the cylindrical column (Smith, 2012).
33 |
Virginia Tech | Smith (2012) identified that the method is ineffective due to the poor recoveries. Later,
instead of dispersing emulsions, spherical agglomerates were tested using the same method.
After several attempts and modifications, the method was successful. Although the dispersion of
agglomerates with an ultrasonic probe was effective, several operational issues were observed.
Since, the viability of the process involving ultrasonic energy was evaluated during the initial
phase of the current research work; the details of the process are discussed in Chapter 2.
The aforementioned research activities from the past decade at Virginia Tech have played
a significant role in determining the key factors in the development of the innovative combined
cleaning and dewatering technology. From a thermodynamic point of view, the concept of
hydrophobic displacement of surface water can produce product moisture at a level that can only
be achieved by thermal drying. None of the methods explored previously can be scaled-up safely
and economically, therefore, the biggest challenge is to develop a well-engineered system to
demonstrate the concept of dewatering-by-displacement on a large-scale, which is necessary to
commercialize this technology. This research, as described in further chapters, particularly the
development and scale-up of a low energy mechanical device to break the agglomerates, has
been a huge step forward in achieving this goal.
In the later stage of the research, a commercial name given to the proposed novel
cleaning and dewatering method — called the Hydrophobic-Hydrophilic Separation (HHS)
process. In later chapters, the innovative method will be referred with the new name.
1.8 Research Objectives
The objectives of this research are:
To fully develop a well-engineered Proof-of-Concept (POC) pilot-scale plant for the
innovative cleaning and dewatering technology that can be employed commercially to
recover the finest coal particles that are now discarded due to their high moisture content.
This goal has been achieved by:
o Conducting the fundamental studies and developing a bench-scale low energy
mechanical device for breaking the agglomerates.
o Conducting comprehensive batch-scale testing with the novel breaking device and
defining the parameters governing the process that were used to scale-up the
process.
34 |
Virginia Tech | o Developing a continuous bench-scale process to enable the design of mass-water-
pentane balanced flowsheet that can be used for the selection of equipment for the
construction of a POC pilot plant with a capacity 100 pounds/hour feed.
o Constructing and scaling–up of the novel mechanical breaking device for the
required capacity.
o Conducting pilot-scale tests with several types of fine coal slurries with the newly
constructed POC pilot plant to demonstrate the cleaning and dewatering
capabilities of the innovative technology.
o Establishing engineering criteria and determining the process economics for the
design and operation of an industrial demonstration plant that will be constructed
by the project sponsor.
To demonstrate the capability of the innovative process for reducing environmental
impacts associated with the fine coal slurries while simultaneously creating a potential
source of new revenue and profit for coal producers around the world.
1.9 Research Organization
This dissertation consists of seven chapters. The first chapter has provided detailed
background information and a comprehensive review of the previous investigations that have a
pivotal role in the development of this innovative technology. The second chapter discusses the
batch-scale testing program and the development of a laboratory-scale mechanical device (the
heart of this process), batch-scale test result and modifications that assisted in the development
of a process engineering flowsheet for the POC pilot plant. The third chapter discusses the
fundamental aspects and the scientific studies conducted for understanding the proposed
technology. The fourth chapter illustrates the construction and engineering of a proof-of-concept
(POC) pilot plant. In addition, shakedown testing with newly constructed POC plant and the
preliminary pilot-test results are discussed. While the fifth chapter provides the complete pilot-
scale testing program with results and evaluation of the POC pilot plant, the sixth chapter
analyzes the different engineering models for each unit operation involved in the process. The
models and analysis can be valuable in designing the demonstration plant for the sponsors.
Finally, the seventh chapter summarizes the whole research work and proposed future
recommendations that can help in improving the process.
35 |
Virginia Tech | CHAPTER 2 – Batch-Scale Developments of the HHS Process
2.1 Introduction
In the coal preparation plants, froth flotation is the most recognized technique used for
cleaning fine (150 microns x 0) coal (Osborne, 1988). Flotation is a water-based separation
process, which requires industry to use a dewatering step to produce sellable products. The
flotation concentrate (clean coal) is typically dewatered by conventional means such as vacuum
filters, screenbowl centrifuges, or by advanced dewatering methods like the hyperbaric
centrifuge technology (Schultz et al., 2012). Nonetheless, existing technologies cannot produce
single-digit moisture values that can replace thermal dryers, which are very expensive (Osborne,
1988) and no longer considered as a viable drying option due to regulations and several
restrictions in the United States.
Another fine coal cleaning process that has been explored in the past is selective
agglomeration. Several studies conducted on the oil agglomeration process using fine coal feeds
and short chain hydrocarbons achieved better combustible recoveries (Keller, 1985) compared to
conventional flotation, but the process was not preferred in the United States due to high costs
associated with oil consumption. Furthermore, typical moisture in the coal agglomerates is 40%
by weight (Keller, 1985), which makes the product an undesirable commodity in the current
market.
In light of this, researchers at Virginia Tech developed an innovative technology
involving the concept of dewatering by displacement, which was first proposed by Yoon and
Luttrell (1995). Since surface forces get stronger for micron size particles, an additional
proprietary separation stage was added to treat the agglomerated coal. The novel technique
simultaneously cleans and dewaters well-liberated fine coal feedstocks (some of which are
currently discarded) and provides a final product with low single-digit ash and moisture contents
at a very low energy.
The proposed technology is called the Hydrophobic-Hydrophilic Separation (HHS)
process. The theoretical concept of hydrophobic-hydrophilic separation was tested on a bench-
scale reactor using several coal-feed samples from the United States to demonstrate the
thermodynamics behind the proposed process. The bench-scale unit was specifically designed to
40 |
Virginia Tech | serve the purpose of identifying and evaluating key process parameters, which later will be
helpful in the scale-up of the process. This chapter discusses the development of the HHS
process bench-scale systems, designing of a novel low-energy dispersion device, and engineering
analysis of batch-scale testing data for the development of the process flowsheet.
2.2 Experimental Procedures
This section describes the HHS process laboratory testing by various methods utilizing
non-polar hydrophobic liquids to separate both moisture and mineral matter impurities from fine
coal samples. Several challenges were encountered during the development of batch-scale testing
process and, therefore, several modifications to the process were required, which will be
discussed in later sections. Two methods for breaking the agglomerates, including the use of a
novel vibrating mesh design, were evaluated using fine coal samples from major coal preparation
plants in the United States.
2.2.1 Material and Method
During the course of development and testing, the only hydrophobic reagent used was n-
pentane liquid. Pentane, because of its low boiling point (98°F), is easy to evaporate and recover
by condensation. Liquid pentane (C H ) is a colorless, immiscible liquid and short-chain
5 12
aliphatic hydrocarbon with an alkyl radical group, which is hydrophobic in nature. The density of
n-pentane is 0.626 g/cm3; therefore pentane liquid floats on top of water in a pure pentane-water
system. Table 2.1 outlines some of the physical and chemical properties of n-pentane. For bench
scale experiments, n-pentane (98% pure) was procured from Alfa Aesar.
The experimental set-up included a regular kitchen blender (Black & Decker BLC12650)
equipped with variable speed controller for high/low shear mixing, a 60 mesh sieve for
separating agglomerates from the “dirty” aqueous phase, and custom-made glass columns
designed and manufactured at the Virginia Tech (Figure 2.1). The coal agglomerates retained on
the top of the sieve were poured manually from the top in the glass column, which was initially
filled with n-pentane liquid and water. During the initial course of batch testing, an ultrasonic
probe was initially used. Later, a low-energy vibrating-mesh device was developed for
dispersion of the agglomerates.
41 |
Virginia Tech | minus 44 micron, was taken from Arch’s Cardinal and Leer preparation plants. Another set of
fine coal flotation feed samples was procured from the Kingston plant at Alpha Natural
Resources.
2.3 Initial Testing with Ultrasonic Energy
The batch-scale tests were initially performed with Bailey’s screenbowl main effluent
samples using ultrasonic energy to break the agglomerates. The feed samples contained 41.1%
ash value by weight on a dry basis and contained more than 90% minus 44 micron particles. The
coal samples, when procured, were decanted to obtain feed slurry with a high percent solid.
2.3.1 Study with Emulsions
The feed was first diluted to 6% solids by weight using fresh water. Equal volumes of
hydrophobic liquid and coal slurry were mixed in a 600 mL container using a kitchen blender at
low intensity. The mixing was continued until the phase separation was visibly observed. The
resulting product consisted of two layers. The upper layer, which was a mousse-like thick coal
mass, consisted of coal-water-oil emulsions floating on top of water. The bulk of the water and
hydrophilic mineral matter separated and settled to the bottom of the blender. The clean coal
mass with pentane floating on the aqueous phase was separated using a 60-mesh sieve. It was
observed that the coal mass retained on the screen had large water globules stabilized by fine
coal particles in oil, which appears like a paste.
The paste-like coal mass was than fed into a 1.5-inch diameter glass column equipped
with multiple overflow ports at different heights. The column was initially filled with clear
pentane and water. An ultrasonic probe manufactured by Qsonica (Model: Q700) operating at a
frequency of 20 kHz was mounted at the bottom of the column so that the tip of the probe
remained in the water phase as shown in Figure 2.2. As the probe operated, the water-in-oil
emulsions were broken in a way that water trapped within the emulsion drained out of the
hydrophobic liquid (n-pentane) phase. Three distinct phases existed within the reactor. The
lowest phase in which the tip of the ultrasonic probe was submerged consisted of water and ash.
Over time, a buildup of emulsions formed at the oil/water interface. The uppermost layer was
mostly hydrophobic liquid with dry coal powder dispersed in it. As the top layer from the mixer
43 |
Virginia Tech | Figure 2.2 Experimental setup used for breaking agglomerates using ultrasonic probe
was fed into the column, the hydrophobic liquid and suspended coal particles exited the column
through an overflow port.
After a certain time period, the emulsion layer became too thick for the ultrasonic waves
to effectively break. Eventually, the column filled up with the emulsions and coal began to exit
the column form the tailings port located at the bottom of the reactor. This method was effective
only for initial time period, and after a long operating time (around 20 to 30 minutes), breaking
emulsions with ultrasonic waves became ineffective. The product sample received in the first
few minutes of the test mostly contained clean liquid pentane with a very small amount of coal
(< 0.2% solid concentration), but not enough to conduct any analysis. Therefore, moisture and
ash values could not be determined for these small products. Similar procedures were followed
in later test runs by changing operating parameters such as mixing intensity, reducing feed
percent solids, increasing agglomeration retention time and reducing oil dosage in order to
eliminate the formation of these stable emulsions.
2.3.2 Study with Agglomerates
Due to the low throughput and inability to continuously run the process with emulsions,
other coal and hydrophobic liquid products that could be fed into the separatory column were
explored. It was observed that when the oil dosage was reduced to less than 30% by weight and
44 |
Virginia Tech | the residence time was increased, the fine coal from the feed slurry formed loosely bound
spherical shape coal agglomerates. Later, the retention time was reduced by employing high-
intensity mixing for the first few seconds, followed by low-intensity mixing for one minute. The
spherical agglomerates formed by this procedure were very easy to disperse (as compared to
emulsions) in the pentane column and therefore followed in further batch-scale testing.
Samples from the Bailey preparation plant screenbowl main effluent and Kingston plant
flotation feed were tested with this method. Coal agglomerates were prepared by mixing 600 mL
of slurry (6% solids) and 10 mL of hydrophobic liquid in the variable speed kitchen blender. For
the initial 20 seconds of mixing, the blender was operated at a high speed to ensure a high-shear
mixing environment. With high-shear mixing, micro-size agglomerates were observed. To grow
the agglomerates, the blender was turned down and operated at low shear for another 40 to 60
seconds. The spherical agglomerates formed floated on top of the water phase. The agglomerates
were poured across a 60-mesh screen to remove “dirty” water containing unwanted impurities of
mineral matter. The water fell through the screen, while the agglomerates remained on top. In
addition, no large stable drops of water were observed in this procedure. The loosely bound
spherical agglomerates were large, usually with top sizes in the range of 0.8-0.9 mm, and had
fairly low moisture values (35 – 55% by weight). By changing mixing time and hydrophobic
liquid dosage, agglomerate size and moisture varied as well.
The ultrasonic probe described in the previous process was used to break the
agglomerates. The column was filled with a small volume of water so that the water level was
approximately 1 inch above the probe tip. The remainder of the column was filled with pentane
up to the overflow port. Then, the coal agglomerates were poured into the hydrophobic liquid
phase. The agglomerates broke up and coal particles dispersed into the hydrophobic liquid phase
almost immediately. Additional hydrophobic liquid was pumped continuously from the top of
the column for continuous flushing of the dispersed solids. The overflow was collected in a
beaker and the hydrophobic liquid was evaporated at 40°C, leaving behind the dry clean coal
product. Due to a higher product solids content (1-2% solids), the throughput with this method
was much higher than the emulsion method. In addition, no build-up of stable emulsions at the
oil/water interface was observed during the test runs. Unfortunately, a major issue was observed
after a long operation time. Due to excessive heat caused by the large energy input, the probe
started boiling the water.
45 |
Virginia Tech | Table 2.2 Batch-scale test results with agglomerates using ultrasonic probe
Kingston Coal, Beckley, Alpha Natural Resources
Feed Ash % Product Product Reject Combustible
Moisture% Ash % Ash % Recovery %
1.19 4.53 92.00 89.95
1.16 4.38 92.44 90.54
54.52
1.07 4.87 90.56 87.97
1.10 4.05 91.67 89.45
Bailey Coal, Consol Energy
1.15 3.06 86.94 89.94
8.42 2.43 84.19 87.30
41.11
0.60 3.36 86.97 90.01
1.01 3.60 87.41 90.43
phase inside the reactor. Therefore, the tests were only conducted for a small time period, usually
less than 10-15 minutes. The test was repeated multiple times with both the feed samples. The
clean coal and tail samples were collected and analyzed for each test. Table 2.2 provides a
complete set of test results. The achieved moisture of the final clean coal product was below 2%
by weight and combustible recoveries as high as 90% were obtained, indicating the process can
separate unwanted components from coal using the concept of hydrophobic displacement.
2.3.3 Development of Semi-Continuous Bench-Scale System
The detailed schematic of the newly constructed semi-continuous bench-scale system is
shown in Figure 2.3. The whole apparatus was designed and assembled using in-house facilities
at Virginia Tech. The continuous bench-scale system has five major units: a mixing vessel, a
phase separator, a settling vessel, an evaporator and condenser, and a reagent recycling vessel.
The preliminary batch-scale testing provided crucial information, such as the process is
ineffective with emulsions and very effective with agglomerates. This is because the coal-
hydrocarbon liquid mixture produces a very stable emulsion with high pentane dosage in strong
mixing conditions. Trial-and-error testing indicated that the stable emulsion could not be broken
with simple agitation, or sound waves, or with high frequency ultrasonic waves, in the separatory
column (phase separator). Spherical agglomerates were prepared with two stages of mixing and a
low dosage of pentane. In the existing system, high-shear mixing was hard to implement with the
small glass units, therefore the mixing chamber was only used for low-shear mixing. The high-
shear mixing was achieved with a kitchen blender and pumped into the mixing vessel.
46 |
Virginia Tech | The phase separator was assembled with a high-frequency ultrasonic probe at the bottom
of the vessel. To overcome the issue of excessive heating, the test cell was equipped with a
cooling water jacket. The overflow from the mixing vessel, which contains coal agglomerates
and the hydrocarbon liquid, creates an interface of water and hydrocarbon liquid in the phase
separator reactor. The ultrasonic energy from the aqueous phase was used to break and disperse
the coal agglomerates. The dispersed particles moved into the pentane phase and eventually
released water and associated ash-bearing minerals into the aqueous phase. Phase separator
columns of different height were used to determine the optimum pentane column height. It was
observed that low column heights worked better. The poorer performance may be due to the
lower energy per unit volume associated with the taller columns.
The biggest challenge faced during the operation was to keep the water-oil interface level
constant in the phase separator. Due to the low column height, the interface level was hard to
maintain, which led to the formation of stable emulsions produced by the ultrasonic energy
inside the reactor. Furthermore, no screen was employed in the existing system after the high/low
shear mixing vessel, because screening was hard to implement in the small-scale continuous
circuit. This also promoted the formation of stable emulsions, as lots of water from the mixing
vessel reported with the agglomerates. Because of these operating issues, further test runs with
this approach were discontinued. In later test runs, the tailing from the phase separator was
pumped constantly at high speed so that the interface height could be controlled.
Once the dispersion process in the phase-separator reactor appeared to work, the clean
coal in the bulk hydrocarbon liquid from the phase separator was pumped to a settler vessel,
where coal settled with time and was transferred into the evaporator, while liquid hydrocarbon
reported back to a reagent tank from overflow of the settling vessel. The evaporator used was a
closed glass jar with vapor ports, which was placed in a hot water container mounted on a hot
plate. The temperature inside the glass jar was maintained between 40°C to 50°C. The
hydrocarbon from the settler inside the jar eventually boiled out and was captured using a two-
stage condenser that passed the pentane back into the reagent tank. The condenser was equipped
with a water chiller unit that maintained the cooling water at 5°C.
47 |
Virginia Tech | 2.3.4 Dispersion with Ultrasonic Energy – Discussion
Using an ultrasonic probe for breaking emulsions proved to be ineffective for the process.
The maximum overflow concentration of product achieved was 0.2% solids by weight. The low
quantity of product sample made it impossible to analyze for moisture and ash contents. It was
observed that the ultrasonic vibrations did not actually break the emulsions but rather split the
emulsions into smaller and smaller emulsions. It must be noted that ultrasonic energy is a high-
energy source, which is only concentrated close to the probe tip rather than uniformly distributed
in the whole pentane column.
The water-in-oil emulsions formed in the method were very stable in nature, which
makes it harder to break them. It was observed that, due to the high dosage of hydrophobic
liquid, water droplets stabilized by coal particles were found suspended in the oil phase. It is well
known that hydrophobic particles, such as micron-size coal, with three-phase contact angles
greater than 90o can act as “particle surfactants.” Therefore, these particles can stabilize water
droplets in the bulk oil phase and form water-in-oil emulsions (Binks, 2003). These emulsions,
which are highly stable in nature, contain 60 to 80% water and resemble “chocolate mousse”
(Fingas and Fieldhouse, 2004).
On the contrary, the fine coal spherical agglomerates typically have 40% moisture by
weight (Keller, 1985) and do not contain stable water globules, as was found in emulsions.
Breaking agglomerates using ultrasonic vibrations was successful both in cleaning and
dewatering coal. The main advantage of using agglomerates over emulsions was a higher
throughput in the product. When agglomerates were introduced in the reactor, they quickly
dispersed into the pentane column. Consistent low moistures and high recoveries were achieved
with the dispersion of agglomerates.
While the breaking of agglomerates with ultrasonic vibrations was successful, a major
operating issue was encountered in using the probe, i.e. heating of the pentane and water column.
The ultrasonic probe operated at a frequency of 20 kHz, which caused the tip to heat excessively.
The probe was always placed in the water phase to isolate the pentane phase from the ultrasonic
tip. However, after approximately 15 minutes of operation, the probe would generate enough
heat to boil the water, which in turn caused the pentane layer to boil. After operating for
approximately 20 minutes, large cavities started appearing at the tip of the probe, and the glass
column was very hot when touched. It is believed that the water directly against the tip was hot
49 |
Virginia Tech | enough to boil, causing these cavities. At that point, the tip of the probe was very hot and was
considered to be unsafe for further operation. Therefore, experiments were never conducted for
longer than 25 minutes. Even though the results were highly impressive, this issue generated a
serious threat in scaling-up the technology.
Another problem was the cost associated with scaling up of ultrasonic source of energy.
It is not economical considering the market value of the product. Due to the severe issues with
safety and scale-up of the ultrasonic technology, other possible designs for breaking the
agglomerates were explored that could be more readily utilized on a commercialized scale.
2.4 Development of Batch-Scale Vibrating Mesh
Spherical agglomerates are held by intermolecular forces (Kendall, 1988), which are
weak range forces. Interfacial thermodynamics of the process, which will be discussed in detail
later in this document, showed that the high amount of energy provided may overcome free
energy associated with cohesion of water droplets released after agglomerates dispersion, and
therefore can hinder their coalescence mechanism. The thermodynamics also showed that the
energy requirement is very low for a pentane liquid to displace a water drop from the coal
surface in the three-phase mixture. In light of this, a simple mechanical device was explored that
could serve the purpose of effectively breaking of agglomerates, keep the dispersed particle in
suspension, and accelerate the process of coalescence of water droplets.
Richardson and Thorpe (1995) developed a simple mechanical device for dispersion that
is used in the dairy industry. The apparatus was designed to measure milk coagulation time and
rigidity in formation of fermented dairy products. This apparatus included a flat disc-shaped
probe that was suspended on a wire placed into a fermented dairy product-making vessel filled
with milk. The probe was reciprocated through a small vertical distance within the coagulating
milk in the vessel. This mechanical device was installed on the top of the tank with the disc
completely submerged in the milk and operated at constant low frequency below 2 Hz
(Richardson and Thorpe, 1995).
This Richardson and Thorpe (1995) concept was modified and implemented in the
current work for the purpose of breaking agglomerates and promoting water coalescence in the
pentane column. Similar to the abovementioned device, where particles of milk fat find each
other and coagulate, the developed mechanical vibrating device may assist the released water
50 |
Virginia Tech | Figure 2.4 Constructed vibrating mesh lab scale unit.
droplets in finding each other and coalescing. T he shaker was designed only to create up-down
motion. A thin shaft with two variable size mesh discs of aperture 0.5 mm and 80 micrometer
was connected to an electro-dynamic shaker, manufactured by Modal Shop INC model 2007E.
Considering that frequency and amplitude of vibration are important parameters in controlling
vibrating energy, the shaker was equipped with a variable speed controller device, which
provided frequencies ranging from 0 – 60 Hz at variable amplitudes. The discs diameter was kept
close to the inside diameter of the reactor to provide maximum surface area for dispersion. The
two meshes (0.5 and 0.08 mm aperture) were separated one inch apart with the bottom disc
connected at the end of the shaft. The complete assembly is exhibited in Figure 2.4.
2.4.1 Experimental Procedure with Vibrating Mesh
Prior to dispersion, similar methods were used to form the coal agglomerates utilizing the
kitchen blender. The separation took place in a custom-made glass reactor that was 5-inches high
and 1.5-inches in diameter. The reactor was initially filled with water only up to one-third of the
total height. The newly developed vibrating mesh device was inserted inside the reactor from the
open top such that the lower disc was just above the water surface. The reactor was then filled
51 |
Virginia Tech | with pentane liquid just below the overflow port. The schematic of the laboratory testing system
is shown in Figure 2.5.
The electro-dynamic shaker, which controls the vibrating mesh device, was operated at a
constant frequency throughout the testing period. Agglomerates were then poured from the top of
the reactor. The fine coal particles from agglomerates started dispersing almost immediately (as
visually observed) in the pentane column leaving the residual water and mineral matter in the
aqueous phase. The dry coal in pentane liquid was collected from the overflow port of the reactor
in a glass beaker. The evaporator-condenser unit, which was designed for the continuous bench-
scale system, was used to evaporate and recover the pentane consumed in the process. The
pentane-free clean coal product was collected manually from the evaporator glass vessel and
analyzed.
The novel mechanical vibrating-mesh device worked very efficiently as the analysis
results showed similar moisture and recoveries as obtained using the ultrasonic probe. The
biggest advantage with this system is that it is very safe to use. The method also utilizes only a
small amount of energy and produced very high quality products. Also, due to the inherent
simplicity of the vibrator, the device can be scaled-up. Another crucial observation was that this
method provided a much higher throughput due to the higher solids content of the overflow
product (up to 12% solids). It is believed that the vibrating device also creates a uniform energy
distribution in the reactor, unlike the ultrasonic probe.
2.5 Batch Scale Results – Mechanical Vibrating Mesh
Breaking agglomerates with the novel low-energy mechanical vibrating mesh was used in
both cleaning and dewatering fine coal material. The dispersion rate of the fine particles was
observed to be higher compared to previous methods, which resulted in higher throughputs and
lower moisture clean coal products. Several coal samples were tested, from metallurgical coal to
steam coal, as well as particle size ranges from minus 325 mesh (screenbowl main effluent and
6-inch diameter “deslime” cyclone overflow) to minus 100 mesh (flotation feed). Results from
the testing of each coal sample are discussed separately in this section.
52 |
Virginia Tech | Figure 2.5 Constructed bench-scale experimental set up for testing vibrating device
2.5.1 Screenbowl Main Effluent
Screenbowl centrifuges are commonly used to dewater flotation products in coal
preparation plants. The overflow of the centrifu ge bowl, which is the main effluent, is rejected
into thickeners as waste. It is well known that the effluent loses almost 50% of the ultrafine
(minus 325 mesh) particles from the flotation concentrate (Luttrell, 2011). At the preparation
plants, this stream usually has flotation chemicals, such as frothers and collectors. Also, it has
very low percent solids by weight, typically 3-6 % in range.
Four different screenbowl main effluent samples were tested — two each from Consol
Energy and Arch Coal. The samples were procured from Consol’s Bailey and Buchanan plants
and from Arch’s Beckley and Sentinel plants. Tables 2.4 – 2.7 summarize the test results of
screenbowl main effluent using the HHS process bench-scale system. As shown, the process
produced consistently low moistures and high coal recoveries. The combustible recovery on
metallurgical coals, such as those from the Buchanan, Beckley and Sentinel plants, was
exceptionally high (96 – 99%) in all the batch-tests, while the steam coal recovery from Bailey
plant was as high as 90%. As anticipated, the moisture percentage in the final product was in
single digits for all types of coal, indicating the HHS process is ideally suited to recover lost coal
in discarded streams from preparation plants.
53 |
Virginia Tech | Table 2.8 Batch test results from Cardinal 6” deslime cyclone overflow
Feed Ash % Product Product Reject Combustible
Moisture% Ash % Ash % Recovery %
3.1 2.8 84.2 78.8
3.5 3.9 88.0 84.7
3.8 2.9 87.2 83.4
53.6
10.6 3.0 85.0 80.1
9.1 3.7 89.4 86.7
Table 2.9 Batch test results from Leer 6” deslime cyclone overflow
Feed Ash % Product Product Reject Combustible
Moisture% Ash % Ash % Recovery %
3.4 3.6 89.3 87.9
50.9
6.1 4.5 87.3 85.5
tailings thickeners. The percent solids in deslime cyclone overflows are typically higher (6-8%)
co mpared to screenbowl main effluents. In addition, since it is a run-of-mine raw feed, the
stream is completely free of chemicals.
Two deslime cyclone overflow samples were tested, both from Arch Coal. The samples
were procured from the Cardinal and Leer preparation plants, both of which process high-grade
bituminous coals that are sold into the premium metallurgical coal market.
The results shown in Tables 2.8 – 2.9 indicate that the HHS process can achieve high
combustible recoveries in range from 79-88% with product moistures between 3-10%. The HHS
process also responded very well with the deslime cyclone overflow feed and, therefore, can be
used to recover this process stream.
2.5.3 Flotation Feed
The objective of this portion of the research study was to develop and demonstrate the
HHS process for recovering coal from ultrafine discarded streams. However, the scope of this
technology is not limited, and the ultimate goal would be to modify the existing fine coal circuit
by replacing the flotation process with the HHS technology, provided it responds well to
upgrading of flotation feeds.
In light of this, two different flotation feed samples were tested with HHS process. The
feed size is typically composed of 100 x 325 mesh particles. Samples were procured from
55 |
Virginia Tech | Table 2.10 Batch test results from Bailey flotation feed samples
Feed Ash % Product Product Reject Combustible
Moisture% Ash % Ash % Recovery %
1.1 5.5 87.1 88.7
0.7 4.6 87.7 89.2
2.7 3.0 88.7 90.0
1.5 3.1 88.5 89.8
44.9
3.2 3.4 87.9 89.2
4.7 3.2 89.2 90.5
1.6 3.4 89.1 90.4
1.0 3.5 89.1 90.4
55.7 1.8 4.7 90.6 87.4
Table 2.11 Batch test results from Kingston flotation feed samples
Feed Ash % Product Product Reject Combustible
Moisture% Ash % Ash % Recovery %
0.7 3.2 91.6 90.1
52.6 1.0 3.6 91.7 90.3
0.7 3.5 91.4 89.9
51.0 1.1 4.2 90.0 88.9
the Bailey Preparation Plant, which owned by Consol Energy and is located in the northern
Appalachian Coalfields, and the Kingston preparation plant which owned by Alpha Natural
Resources, and is located in the central Appalachian Coalfields.
Batch test results of flotation feed tested with the novel technology bench-scale system
consistently achieved exceptionally low moisture and high combustible products, as exhibited in
Tables 2.10 – 2.11. The combustible (carbon) recovery on Bailey’s and Kingston’s coal samples
were as high as 90.4% and 90.1%, while the moisture percent was in range of 0.7-4.7% and 0.7-
1.0%, respectively.
2.5.4 General Observations
The low energy mechanical vibrating mesh was proven to be very successful in breaking
the agglomerates during the batch-scale testing. This method can be scaled up and is economical
and safe to use as compared to the alternative approach of ultrasonic vibration. This innovative
breaking device creates a uniform hydrodynamic shear field in the hydrophobic reagent column,
which effectively disperses the coal particles and keeps them suspended in the column.
Furthermore, the mesh design and low vibration frequency promotes water coalescence in the
hydrophobic liquid phase, which is crucial for moisture reduction.
56 |
Virginia Tech | Figure 2.6 Final tail and product samples produced from HHS process bench-scale system
The vibrating mesh device was proven to be highly effective in producing consistently
low moisture and ash values. In addition, the mesh provided higher percent solids in the
overflow as compared to the ultrasonic device, which considerably increased the throughput
pr oductivity from the process. In the method, no formation of stable water-in-oil emulsions was
observed during any of the test runs. The bench-scale system of HHS process demonstrated
excellent cleaning capabilities with ultrafine particles, as illustrated in Figure 2.6. The
photograph corresponds to final reject samples with ash contents of 80-90% and a final clean
coal product with less than 5% moisture and ash. For an efficient separation, the bottom screen
of the mechanical vibrating mesh device must be just above the water-pentane interface inside
the column, otherwise the formation of micro-emulsions were observed after long periods of
operation.
The major operating parameters that control the input vibration energy were found to be
the vibrational frequency (f) and the amplitude of the vibrations (A). It was observed that these
parameters affect the process efficiency individually and in the combination of each other. In
addition, it was also observed that the ratio of the amplitude of vibration and pentane column
length also affects the product quality. In light of this, parametric studies were conducted
focusing the effect of vibration energy and length ratio on product moisture. The details are
57 |
Virginia Tech | discussed in Chapters 3 and 4, as these two criteria were considered for the development of
vibrating mesh reactor for the HHS process Proof-of-Concept (POC) demonstration plant.
2.6 Bench-Scale Testing Results - Discussion
The results obtained from bench-scale testing of HHS process have successfully
demonstrated that the concept of dewatering by displacement can be implemented for the
efficient cleaning and dewatering of ultrafine particles. To evaluate the cleaning capabilities of
the process, recovery-rejection curves were plotted for all the samples tested on bench-scale, as
illustrated in Figure 2.7. The diagonal lines represent the separation efficiency of the process.
The “separation efficiency” is a performance level indicator based on carbon recoveries and ash
rejection values. Mathematically, the separation efficiency is defined as the recovery of desirable
material in a given product minus the recovery of undesirable material in the same product. In
the case for coal cleaning, the separation efficiency (E) can be obtained from Equation 2.1.
E = R – (100 – J) = R + J - 100 [2.1]
In this equation, R is the combustible recovery and J is the ash rejection. R represents the
percentage of combustible matter present in the feed that reports to the clean coal, while J
represents the percentage of ash present in the feed that reports to the reject.
The performance evaluation of a process using separation efficiency is useful, as both
recovery and rejection terms normalizes the variation in the feed. As shown in Figure 2.7, the
separation efficiency achieved using the HHS process was very high in all the cases, although
some small variations are noted because of differences in the feed ash of each stream. In the case
of the screenbowl main effluent feed samples, which are typically 10-15% ash, almost all of the
carbon from the feed is recovered. In cases involving the deslime cyclone overflow or flotation
feed samples, which are typically much higher in ash, the data showed that almost all the ash
from the feed is discarded in the tailings.
The biggest advantage of the proposed novel process is the moisture reduction associated
with fine particles. Conventional dewatering methods produce moistures between 10-30%, which
is still above most market specifications (typically 6-8% surface moisture). These traditional
methods also require low water content in the feed for efficient dewatering. Screenbowl
centrifuges produce 16-18% moisture product and can accommodate up to 35% water in the feed
(Keles, 2010), but this technology typically loses half of the minus 325 mesh coal particles
58 |
Virginia Tech | Figure 2.7 Separation efficiency data for the HHS process bench-scale system
(Luttrell, 2011). Vacuum disk filters, which were once popular in the United States, produce 22-
25% moisture products from higher moisture feeds and recover up to 97% of the solids
(Osborne, 1988). An advanced technology, hyperbaric centrifugation system, can produce as low
as 13% moisture product from 10% solid feed (Keles, 2010), although few industrial installations
of this technology have been implemented due to the high capital cost of this unit. The laboratory
test results have showed that the HHS process is far superior to any of these existing dewatering
technologies. The bench-scale system has demonstrated that a low moisture product (below
10%) is possible by using HHS process, which in the past could only be achieved by thermal
dryers.
Figure 2.8 depicts the range and average values of clean coal product moisture
(percentage by weight) of each coal sample tested on HHS process bench-scale system. The
moisture achieved is below the target moisture of 10% in almost all of the test runs, which
demonstrates the highly efficient dewatering capabilities of the novel technology.
59 |
Virginia Tech | Abbreviations: FF – Flotation Feed; DCO – Deslime Cyclone Overflow; SBE – Screenbowl Main Effluent
Figure 2.8 Product moisture% ranges with average values obtained from HHS process
bench-scale system
One important factor that was also evaluated during the batch scale testing is the dosage
of pentane in the agglomeration process. Dosa ge increases as the particle size gets finer. For
example, the pentane dosage had to be increased for the screenbowl effluent coal samples up to
25-30% by weight of dry feed to attain success ful spherical agglomeration. This value is much
higher than the 10-15% by weight pentane dosage required for the flotation feed sample. The
reason is due to the increase in the total surface area in the ultrafine particles.
2.7 Hydrophobic Liquid Consumption and Recovery
Consumption or loss of pentane in the process is one of the most crucial aspects for a
successful commercialization of proposed technology, both in terms of economics and
environmental regulations. Although the process produces low moisture and highly clean
product, a high loss in producing a low commodity product like coal can hinder its development.
60 |
Virginia Tech | Data for the solubility of pentane in water (40 mg/L at 20°C) is well known, but no information
for the loss of pentane associated with coal, either via absorption or adsorption, has been
published. In order to estimate the loss associated with clean coal product, basic attempts were
made at a batch-scale to quantify the loss of pentane theoretically as well as experimentally. It is
important to mention that the pentane absorption in coal may depend on many design parameters
and heat exchanger efficiencies and, therefore, can be estimated precisely only at a large scale.
A theoretical model was developed based on the mole concept. Consider the clean fine
coal product collected in a sealed jar of 1 liter volume at standard temperature and pressure
(STP) conditions. Assuming the voids are completely occupied by trapped pentane vapors, a
linear relationship (shown in Figure 2.9) can be established between pounds of pentane in a ton
of fine coal and the void fraction using the ideal gas equation. For example, for a 20% of void
fraction (which is the available volume of pentane gas), relative loss associated with coal product
would be 1.88 lbs/ton of coal at 20°C. Similar predictions can be made for ultrafine coal
particles. The typical bulk density of coal powder is 0.641 g/cm3. Considering the clean coal
particle density 1.25 g/cm3 and with 10% moisture, the available void fraction for pentane vapors
is 42.3%. From the Figure 2.9, the predicted pentane loss would be approximately 3.97 lbs/ton.
Figure 2.9 Empirical analyses for estimation of pentane loss/ton of coal relative to void
fraction
61 |
Virginia Tech | A series of experiments was conducted to evaluate the loss of pentane with ultrafine dry
coal powder. Equal amounts of dry coal powder, in separate sealed glass vials, were completely
soaked with pentane liquid and left overnight. After that, each vial was heated in isothermal
conditions for different time period (ranging from 0 – 30 minutes) and the weight percent gain in
the dry coal samples was recorded with respect to time (at every 3 minutes interval). The gain
was assumed to be due to pentane absorption. The procedure was used to develop pentane
absorption rate curves for five different temperatures (55 - 95°C), as exhibited in Figure 2.10.
The study indicated that the rate of evaporation of pentane increased as the temperature
increased. Clearly the inflection point on these graphs shows two rate constants for evaporation
with a sharp inflection point. Initially, the rate is higher, as the bulk of pentane is easily available
for evaporation at the surface. After crossing the inflection point, the rate decreases drastically as
only a small amount of pentane may be trapped in between the coal particles void space or
absorbed/adsorbed with coal. Table 2.12 summarizes the recorded loss of pentane after 30
minutes at each temperature.
Table 2.12 Estimated pentane loss after 30 minutes from experimental studies at
different temperatures
Pentane Loss
Temperature (°C)
(pounds of pentane/metric ton of coal)
55 4.85
65 3.75
75 2.65
85 1.54
95 0.44
2. 8 Conclusions
The HHS process serves two purposes: cleaning and dewatering of fine coal particles. In
many current coal processing plant circuits, ultrafine particle (44 micron x 0) streams —
screenbowl main effluents and 6-inch deslime cyclone overflow — are not processed and are lost
as waste to tailings thickeners. There is no existing technology available that can clean and
dewater this ultrafine material economically. In addition, flotation is a widely accepted process
for treating fine particles (150 x 44 microns) but is a water-based process. Consequently, the
63 |
Virginia Tech | dewatering cost associated with the flotation product exponentially increases the cost of the final
product as particle size is reduced.
Studies conducted using a batch-scale HHS unit showed that too much or too little energy
input is detrimental to the separation performance. Too little energy results in inadequate
breakup of coal agglomerates, poor dispersion of coal particles into the pentane phase, and slow
coalescence of water droplets in the pentane column. Too much energy results in the formation
of coal-water-pentane emulsions that are very difficult to destroy once formed. In light of these
limitations, a low-energy vibrating mesh device was developed for dispersion of agglomerates
and water coalescence, which performed very effectively. The successful laboratory-scale testing
with various types of fine coal feedstocks showed that the HHS process can fulfill the needs of
coal industry by recovering and dewatering discarded ultrafine coal streams, thus increasing the
productivity of existing preparation plants. Furthermore, attempts were made in determining the
loss of pentane associated with coal, which is highly critical for the successful commercialization
of the process. Data obtained using both theoretical and experimental approaches were found to
be consistent and suggest that the expected losses are within acceptable levels.
References
1. Bethell, P.J. and Barbee, C.J. (2007), “Today’s Coal Preparation Plant: A Global
Perspective”, Designing the Coal Preparation Plant of the Future edited by Arnold, B.,
Klima, M. and Bethell, P., Pages 9-20, Society for Mining, Metallurgy, and Exploration.
2. Binks, B.P. (2003), “Emulsions Stablized Solely by Colloidal Particles”, Advances in
Colloid and Interface Science, Volume 100-102, Pages 503-546, Elsevier.
3. Fingas, M. and Fieldhouse, B. (2004), “Formation of Water-in-Oil Emulsions and
Application to Oil Spill Modeling”, Journal of Hazardous Material, Volume - 107, Pages
37-50.
4. Keles, S. (2010), “Fine Coal Dewatering Using Hyperbaric Filter Centrifugation” PhD
Dissertation, Mining & Minerals Engineering, Virginia Tech.
5. Keller, D.V., Jr. (1985), “Agglomeration Type Coal Recovery Processes”, Canadian
Patents Number: CA 1198704.
6. Keller, D.V., Jr. and Burry, W. (1987) “An Investigation of a Separation Process
involving Liquid- Water-Coal Systems,” Colloids and Surfaces, Amsterdam, Volume -
22, Pages 37-50.
7. Kendall, K. (1988), “Agglomerate Strength”, Powder Metallurgy, Volume 31, Page 28.
8. Luttrell, G.H. (2011), [Personnel Communications]
64 |
Virginia Tech | CHAPTER 3 – Fundamental Studies for the HHS process
3.1 Introduction
While the fine coal fraction (minus 150-micron particles) is often recovered in coal
preparation plants using froth flotation, an emerging practice is to remove and discard the
ultrafine fraction (minus 44 micron) from the flotation feed because of associated low recoveries
and high dewatering costs (Bethell and Barbee, 2007). In light of this, researchers at Virginia
Tech have developed an innovative technology, called the HHS (Hydrophobic-Hydrophilic
Separation) Process, utilizing the concept of hydrophobic displacement to simultaneously
recover and dewater well-liberated ultrafine particles. The concept of hydrophobic displacement
of water from a hydrophobic coal surface in the presence of hydrophobic liquid is a
thermodynamically favorable process. The condition for this displacement to occur is based on
the three-phase equilibrium contact angle, which must be greater than 90°.
The HHS technology involves the selective oil agglomeration process as an initial step,
combined dewatering-cleaning (the novelty in the process) as a second step, and oil recovery as
the third step. The successful development of a bench-scale system for the process has proven
that the HHS concept is feasible at a small scale by providing consistent low-moisture high-
quality coal products from variable coal feedstocks. As for any new process, scientific studies are
equally important as the engineering data in the further development of the technology. These
investigations provide a better understanding of the mechanisms that control the process.
Several scientific studies with variable coal types (Capes et al, 1974; Keller, 1981; Capes
and Germain, 1982; Wheelock, 1982; Slaghuis and Ferreira, 1987; Drzymala et al, 1988;
Skarvelakis et al, 2006) can be found on the oil agglomeration process, but the focus of these
studies is mostly on optimizing the coal cleaning (recovery) aspects and very little on moisture
reduction (Capes and Germain, 1982; Smith, 2008). In a standard coal agglomeration method,
which uses low chain recoverable non-polar hydrocarbons, product moistures are typically
reduced to 40% by weight (Keller, 1985). As dewatering of ultrafine coal slurry is an integral
part of the proposed technology, this chapter outlines scientific investigations conducted to
identify two key mechanisms:
66 |
Virginia Tech | how moisture is trapped in an agglomerate structure, and
how moisture is released in the unique dewatering step of HHS process.
Based on these investigations, a theoretical model for the HHS process is proposed that is
supported by the thermodynamics of the system.
3.2 Agglomerates Characteristics
The three crucial governing factors that affect the oil agglomeration process (Capes and
Darkovich, 1984) are:
free energy relationships at the three phase (oil-water-solid) interface,
dosage of an agglomerant (bridging liquid) with respect to carbon content in the feed, and
mixing conditions such as time, intensity and method of mixing.
The free energy relationships for the pentane-water-coal system have been discussed earlier in
Chapter 1. It was identified that the liquid pentane makes a 106° equilibrium contact angle in the
three-phase system and, therefore, quickly wets the hydrophobic coal surface. The other factors
are explored here particularly to study their effect on agglomerate structure using liquid pentane,
which is important to determining how moisture gets trapped in the agglomerates of ultrafine
coal particles.
3.2.1 Agglomerate Structure
Typically, there are three types of structures that can form depending on the amount of oil
used during the formation of agglomerates. These are pendular type, funicular type, and capillary
type (Capes and Jonasson, 1989). An agglomeration study was conducted with mono-size (75
micron) hydrophobized silica particles under a microscope. Figure 3.1, taken with the camera of
the microscope, illustrates with solid particles formed floc-like structures in a pendular form
when using small amounts of the bridging liquid. With higher additions of bridging liquid, these
pendular agglomerates consolidated into more compact funicular structures. Finally, with high
quantities of bridging liquid, the agglomerates become are more compact, like pellets where the
bridging liquid is in the capillary state and the aqueous phase is left out from the agglomerate
structure. An excessive dosage of oil resulted into a formation of highly stable water-in-oil
emulsions, which resembles a mousse in terms of consistence.
67 |
Virginia Tech | Figure 3.1 Image showing types of agglomerate formed with 75 micron silica particles
3.2.2 Entrapment of Moisture in Agglomerate Structure
Water can be trapped in any agglomerate structure by three possible ways, as exhibited in
Figure 3.2. First, the bulk of water is associated with water droplets that are stabilized by micro-
agglomerates. Second, a good proportion of mo isture is trapped in the void spaces inside the
micro-agglomerate structure and finally, by nature, water likes to be attached on any hydrophilic
site available on coal particle surface. For ultrafine coal particles (minus 44 microns), the latter is
expected to be the least likely scenario, as coal and clay particles are usually well liberated for
ultrafine particles.
In order to better identify the dominant mechanism responsible for water entrapment,
several microscopic investigations were conducted with fine coal particles (150 micron) in the
presence of a bridging liquid (pentane). A dilute slurry (5% solids by weight) containing fine
coal was prepared and agglomerates were formed by adding a small dosage (5% by weight) of
pentane. The agglomerates were screened and analyzed under the microscope. As both water and
pentane are clear liquids, it was hard to make a distinction between the two phases. After few
minutes, due to the radiant heat of microscope bulb, the pentane liquid starts expanding, enough
to clearly identify the oil bridges between the particles. The micro-photograph, shown in Figure
68 |
Virginia Tech | 3.3(a), was taken as soon as the pentane started expanding, which eventually evaporated in the
atmosphere. For better moisture assessment, the same experiment was repeated with a high
dosage (30% by weight) of pentane, and a small amount of fluorescein was mixed with the
slurry. The fluorescein is soluble in water and insoluble with pentane. Once the agglomerate
formed, rather than screening, a small portion of the floating agglomerate was sucked into a glass
tube and examined under the microscope. The photograph, shown in Figure 3.3(b), of floating
agglomerates on top of the water surface clearly shows the entrapment of fluorescein dyed water
in fine coal agglomerate.
The preferred size fraction processed in the HHS technology is ultrafine minus 44 micron
particles. Due to limitations with microscope magnification, similar images could not be
generated for this particle size range. To investigate moisture entrapment in agglomerates formed
using ultrafine coal particles, the coal slurry was mixed with fluorescein and the formed
agglomerates were screened out from the bulk water. The agglomerates were collected on a glass
plate and examined under the microscope. Figure 3.4(a) shows a photograph of agglomerates,
which appear to be dry. These agglomerates were then squeezed between the two glass plates
and a photograph was taken as shown in Figure 3.4(b). The image clearly indicates a good
proportion of fluorescein dyed water droplet releasing from the agglomerate structure when the
agglomerate was mechanically squeezed between the two glass plates.
1. Water droplet trapped by micro-agglomerates
2. Water trapped in micro-agglomerate structure
3. Water attached to hydrophilic sites on particle
2
surface
1
3
Figure 3.2 Schematics showing entrapment of moisture in agglomerates
69 |
Virginia Tech | Figure 3.5 Effect of oil dosage on entrapment of moisture in agglomerates
different oil dosages, indicate a significant decr ease in agglomerate moisture as the oil dosage
was increased up to 50% by weight of solids. Dosages higher than 50% did not contribute further
re duction of moisture from the agglomerates, suggesting that all available void space was filled
at this dosage level.
3.2.4 Impact of Mixing Time and Intensity on Agglomerates
The mixing time and intensity of agitation play a very important role in the oil
agglomeration process, particularly in defining the size distribution of agglomerates. An
investigation was conducted with ultrafine coal agglomerates formed by varying impeller speed
at a fixed retention time and vice versa. The agglomerates were screened and a small sample was
spread on a clear white sheet. Photographs were taken utilizing a macro lens from a digital
camera for each scenario and analyzed using image processing software (IMAGE-J) to determine
the size distribution of agglomerates, as depicted in Figure 3.6 and 3.7.
71 |
Virginia Tech | At 2000 RPM and 4 minutes of mixing time, agglomerates were large, compact and
distributed uniformly. When the impeller speed was reduced, the intensity was not enough to
disperse the oil completely and engulf particles. A wide range of size distribution was observed
with a high percentage of micro-agglomerates. On the contrary, at a constant 2000 RPM, but
with a variable mixing time, agglomerates were observed more uniformly distributed even at
short agglomeration times.
The data also indicated that the agglomerate size distributions implicitly defined the
amount of moisture trapped in agglomerates pore volume. A uniform size distribution, as found
with high agitation speed, provided a high percentage of coarser (up to 3 mm) and more compact
agglomerates. This tight structure resulted in a small amount of trapped moisture as exhibited in
the trends shown in Figure 3.8. However, a wide distribution with high percentage of micro-
agglomerates (less than 0.5mm), which were generated with low agitation speed, resulted in a
large amount of water entrapment in agglomerates. Even though the available pore volume in the
micro-agglomerates is much lower, it was observed that some of these agglomerates were
stabilizing large water globules, which might have caused increase in moisture.
Figure 3.8 Moisture entrapment with variable residence time and impeller speed
73 |
Virginia Tech | 3.3 Impact of HHS Process – A Comparison with Agglomeration
The basic difference between the HHS Process and the conventional oil agglomeration
process is an additional stage of coal dispersion/water coalescence after the standard
agglomeration step. This novel step is based on dispersion mechanism, i.e. the breakup of the
agglomerates in a hydrophobic liquid. In this step, coal agglomerates are passed into a liquid
pentane column for dispersion. The hydrophobic dispersed coal particles stay in pentane, while
the trapped moisture from the agglomerates is released and settles to the bottom of pentane
column. Hydrophilic ash-bearing minerals, such as clay, also like to be with the water and are
rejected with the moisture. Thus, this novel step serves a dual purpose of cleaning and
dewatering the ultrafine coal particles.
Figure 3.9 clearly depicts the advantages of the additional step in producing premium
quality coal. In the study, two sets of agglomerates were prepared from a deslime cyclone
overflow reject stream at different dosages of pentane. One of the two sets of samples was
subjected to the novel step of coal dispersion / water coalescence, while the second otherwise
identical sample was handled as in any standard oil agglomeration test. The samples were
analyzed and the results indicate a significant drop in the clean coal product moisture (red circles
vs. blue circles) and an increase in the combustible recovery (red squares vs. blue squares) up to
90%.
Figure 3.9 Comparison of HHS Process with Oil-Agglomeration in terms of product quality
74 |
Virginia Tech | 3.4 Dispersion of Agglomerates
The dispersion of agglomerates formed from ultrafine particles in a hydrophobic liquid is
a key part in the innovation of the novel proposed technology, as discussed in aforementioned
study. Therefore, it is necessary to understand the mechanism, particularly in the proposed
process, to identify the critical aspects of the process. More knowledge generated through
scientific studies provides better understanding of the phenomenon occurring in the reactor,
which is vital for successful scale-up of the HHS process.
Dispersion can be regarded as the result of two separate, though simultaneous, processes:
de-agglomeration and homogenization. De-agglomeration is the breaking of aggregates to
smaller agglomerates or individual particles, whereas in homogenization, each component is
redistributed with its parent component to achieve a more homogeneous mass (Patterson and
Kamal, 1974).
To better understand the dispersion process, it is important to have a fundamental
knowledge of the inter-particle forces that lead to agglomeration. The wetting mechanism of
liquid hydrocarbon on the coal surface without water is controlled by weak van der Waals force
(Keller and Burry, 1987). This force becomes significant when the particle size is less than 1
micron. Other stronger forces may include those due to pendular moisture in the interior of
agglomerates, bridging forces at the point of contact when hydrophilic gangue particles are
present in the agglomerates, and mechanical forces from the interlocking of irregular shape
particle (Hartley et al., 1985).
Dispersion due to agitation in a liquid medium is governed by a competition between the
hydrodynamic shear forces acting on the fine coal agglomerates created by agitation in the liquid
medium and the cohesive inter-particle forces holding the agglomerate together. These
hydrodynamic forces can be determined by studying the strength and geometry of the shear
stress field, while the cohesive forces can be evaluated from particle – particle interaction and
packing arrangements of the agglomerates. Detailed studies on these forces are available in the
published literatures (Rumpf, 1962; Kendall, 1988; Israclachvili, 1991; Bika et al., 2001: and
Boyle et al., 2005).
De-agglomeration in the presence of a shear hydrodynamic field can occur in two ways,
either individually or in combination, by rupture and by erosion (Pontente et al., 2002). Rupture
is the spontaneous process of breaking agglomerates into smaller agglomerates due to the
75 |
Virginia Tech | imposed hydrodynamic stress, by collision between two or more agglomerates and by collision
of agglomerates with the dispersion device. Erosion is the removal of primary particles from the
top layer of the agglomerates, mostly when the shear hydrodynamic field exceeds the cohesive
forces of these particles bonded at the agglomerate surface. Both cases involve the creation of
new interfaces between the agglomerate/particle and the liquid medium.
In HHS process, all of the above dispersion theories can define the mechanism partially,
as it is unknown how the released moisture behaves in the system. From an engineering
standpoint, it is necessary to investigate both kinetics and thermodynamics of the process.
Kinetic studies provide information related to the rate of dispersion, whereas thermodynamics
studies help determine the behavior of fine water droplets released after breaking of the
agglomerates.
3.5 Kinetics Studies of Dispersion
In agglomerate dispersion, two mechanisms occur simultaneously. These are breaking of
agglomerates (de-agglomeration) and suspension of particles in hydrophobic liquid phase
(homogenization). For kinetics studies, the rate of homogenization can be determined with
ultrafine dry coal powder poured in the oscillatory dispersion device on the bench-scale system,
as no breaking mechanism is involved with dry powder feed. Similarly, the combined rate for
both mechanisms can be determined with the agglomerates formed utilizing the same coal feed
under otherwise identical operating conditions. Thus, the difference in rates provides the net rate
for de-agglomeration in the reactor.
Figure 3.10 shows the solid concentration ratios in the reactor at a given time for both the
scenarios. In case of dry coal powder, almost all the particles were recovered from the reactor
(shown with red circles). However, when agglomerates were tested under similar conditions,
some of the particles were retained in the reactor. Therefore, to attain the rate associated with
steady-state conditions, a series of kinetic tests was conducted in batches with equal quantities of
agglomerates. When the first batch of agglomerates was poured into the reactor, a substantial
amount of coal remained at the oil-water interface (as shown with green squares) even at a long
residence time. The test was continued until the overflow from the reactor was visually free of
solids. Then, the second batch of agglomerates was poured into the system and kinetic test was
conducted. Again, a small portion of solids remained in the reactor (shown with green triangles).
76 |
Virginia Tech | When the third batch of agglomerates were introduced, almost all the particles were recovered
(shown with green diamonds), which indicates a steady state condition was reached in the
reactor.
The deviation in the rates between steady state dispersion of agglomerates and
homogenization of dry coal powder can be used to determine true rate for de-agglomeration
mechanism. Figure 3.11 depicts the two recovery plots for the mechanism. The individual
recovery curve can be considered as a residence time distribution plot for breaking of coal
agglomerates in the liquid pentane inside the reactor while the cumulative recovery plot defines
the percentage of agglomerates that break in the reactor at any given time.
It is important to mention that the kinetic tests were conducted at conditions considered to
be optimal for the HHS process bench-scale unit. The curves are expected to shift or vary once
the operating parameters are changed in the reactor. Considering the novelty of the process, a
detailed parametric study with a similar procedure can provide a better knowledge of the system
best suited for this technology.
Figure 3.10 Kinetic rate studies for dispersion mechanism in HHS process
77 |
Virginia Tech | Figure 3.11 Kinetics of de-agglomeration mechanism in HHS process
3.6 Thermodynamic Studies – Proposed Theory
The concept of dewatering by displacement was first disclosed by Yoon and Luttrell
(1995). The idea was to exploit the natural a ffinity of hydrophobic particle surface with a
hydrophobic liquid, which makes the process thermodynamically favorable. In previous years,
several studies were conducted at Virginia Tech to develop the concept into an engineering
system that could recover ultrafine coal particles, which are currently discarded. Smith (2012)
introduced an additional stage after a standard agglomeration process for dispersion using an
ultrasonic device, which can be considered as a large development, though it added another
dimension in the process.
One of the key outcomes of the current research is the development of low-energy
vibrating-mesh device for dispersion, which is safe and can be scaled up. The batch-scale reactor
for dispersion is a cylindrical column initially charged with clear water and liquid pentane and
equipped with the vibrating-mesh device such that the bottom mesh is located just above the
pentane-water interface. The vibrating mesh in the pentane column creates a uniform
78 |
Virginia Tech | hydrodynamic shear field inside the reactor that promotes agglomerate breakup, coal dispersion,
and water coalescence.
Agglomerates, when introduced into the two-phase reactor, break in the hydrodynamic
shear field created by vibrating energy either by rupture or by erosion mechanism. When
agglomerates break in the reactor, they constantly generate new coal surfaces and eventually
disperse all parental particles associated with the agglomerate in the pentane column. The kinetic
data indicated that the breakup is a slow process, although thermodynamics also plays a pivotal
role in defining the system.
The study on structure of ultrafine coal agglomerates has showed that the agglomerates
are held together by pendular bridges formed by liquid pentane between the particles and that the
bulk of the moisture is trapped in the structural voids. The combined moisture content is
typically 40% by weight (Keller, 1985), though it can vary depending upon coal rank, oil
dosages, and mixing time and intensity. During the breakup of agglomerates, the moisture
trapped in the pore volume of the agglomerates gets exposed to bulk pentane, which eventually
displaces (releases) the moisture from the particle surface. This phenomenon, which has been
described by Yoon and Luttrell (1995), creates a complex three-phase system inside the reactor.
Thermodynamically, the work per unit area required for separation (defined as a
reversible work) of water (3) from coal surface (1) in the bulk pentane (2) can be calculated from
energy balance and the Young’s equation. From energy balance,
[3.1]
( )
From the work-interfacial tension relationships and Young’s equation, Equation 3.1 can be re-
written as:
( ) [3.2]
( )
Keller and Burry (1987) estimated surface tension between liquid pentane and water (ϒ )
23
is 51.9 mJ/m2. Sohn et al. (1997) estimated the three phase equilibrium contact angle (θ ) is
123
106°. Therefore, from Equation 3.2, the work per unit area required to separate a drop of water
by pentane from the coal surface is calculated to be 37.6 mJ/m2.
The above relationships help to explain how and why moisture is displaced from the coal
surface, but the unknown the behavior of released water droplet, which still exists in the pentane
column, is not well understood. Theoretically, these fine size droplets should report in the
79 |
Virginia Tech | overflow with clean coal and pentane, as density is almost negligible for micron-size species.
However, the consistently low moistures in the clean coal product obtained with the bench-scale
testing suggests otherwise. In light of this, a theory is proposed, which indicates an additional
phenomenon is occurring in the system.
The proposed theory is partially based on the free energy for attachment/detachment
defined for spherical micron-size particles and partially on the free energy of cohesion. The free
energy for attachment/detachment equation is widely used in colloidal sciences (Binks, 2002) to
identify the minimum energy required for a particle of radius “r” at the water/oil interface to
detach into either of the bulk phases. Assuming an ultrafine coal particle to be spherical with
radius “r”, the free energy for detachment at the water-coal-pentane interface can be determined
by Equation 3.3.
( ) [3.3]
The two terms in parentheses are added when particle is detached from water phase and
subtracted when particle is detached from the pentane phase. Furthermore, the free energy of
attachment can be defined as:
[3.4]
Clearly, Equation 3.4 shows the energy for attachment is always negative and, therefore, is
always a thermodynamically favorable process. On the other hand, the free energy for
detachment (Equation 3.3) is always positive, which indicates some external energy is always
required to detach the particle in either of the bulk phases.
Figure 3.12 graphically represents Equation 3.3 for the solid-water-pentane system. It can
be deduced from this plot that particles like to stay at the pentane-water interface, and a specific
amount of energy is needed to move them into either of the bulk phases. As exhibited, the critical
three-phase contact angle is 90°. If a particle is hydrophilic (such as clay) with contact angle
<90°, it is easier for a particle to go into the bulk aqueous phase. For a hydrophobic particles
with contact angle >90°, particles move into the bulk pentane phase with only a very small
amount of energy. The three-phase equilibrium contact angle (θ ) of high rank coal particle is
123
106°. Therefore, the minimum free energy per unit area for detachment of a coal particle from
interface into the bulk pentane phase is calculated to be 27.2 mJ/m2.
80 |
Virginia Tech | Figure 3.12 Graphical representation of free energy for detachment of a particle in
pentane-water-solid system
Figure 3.13 illustrates the proposed phen omenon for dewatering of coal particles in the
pentane column inside the reactor, which leads to consistent low product moisture. The
phenomenon is described in four different stages showing the most likely thermodynamic states.
The theory is based on the information surveyed during bench-scale testing and is very well
supported by thermodynamic calculations at the three-phase interface.
State I outlined in Figure 3.13 (a) can be described as a thermodynamic equilibrium state
where coal particles always like to stay at the oil-water interface. Any external energy below
minimum free energy for detachment will not disturb this equilibrium and water droplet may
remain attached to the coal surface.
When the external energy is equal or higher than the minimum free energy, which is 27.2
mJ/m2 (calculated earlier), the equilibrium will be disturbed and the water droplet will be
detached from the hydrophobic coal surface, as shown in Figure 3.13 (b) as State II. Once the
water droplet is removed from the coal surface by providing enough energy, the released water
droplet can have two thermodynamically probable options. First, it can attach back to any of the
newly exposed coal surfaces generated during de-agglomeration, as the free energy for
81 |
Virginia Tech | attachment will be negative (in this case ΔG /ΔA = -27.2 mJ/m2), which makes it
attach
thermodynamically possible. Second, the fine size water droplet can find another released water
droplet and coalesce together to form a bigger droplet, as State III depicted in Figure 3.13 (c).
The free energy per unit area (ΔG /ΔA = -103.8 mJ/m2) for cohesion of two water droplets
cohesion
in pentane is calculated from Equation 3.5.
[3.5]
Since the latter case has higher negative free energy, the coalescence process of water is
thermodynamically more spontaneous and naturally favorable. Consequently, micron-size water
droplets quickly combine to form bigger drops, which eventually sink into the aqueous phase, as
described in State IV in Figure 3.13(d). This phenomenon also helps to explain why a water
column below the pentane column in the reactor can be beneficial for moisture separation. Large
water droplets find it easier to cross the oil-water interface and be completely removed from the
active part of the reactor.
The free energy for the cohesion of water droplets also indicates that the separation
process should work most efficiently in a specific range on external energy. Too little energy (E
< ΔG ) will not displace the moisture associated on the coal surface that always likes to stay
detach
in a three-phase equilibrium state. Too much energy (E > ΔG ) can prohibit the water
cohesion
coalescence process and may also break the existing droplets into much smaller droplets that
remain in suspension in the bulk pentane phase, which may eventually end up in the final
product.
As mentioned earlier, the proposed theory for the process is a perspective of what is
observed during the laboratory-scale batch testing, but is very well corroborated by
thermodynamic calculations. From a theoretical standpoint, it can be understood that the
hydrodynamic energy density within the pentane column plays an important role in moisture
reduction during dispersion. Two conclusions can be made from this study, which are critical in
designing the reactor:
The reactor should promote the water coalescence mechanism in addition to de-
agglomeration and homogenization mechanism of coal particles.
To completely remove the moisture (<1%) and achieve a dry coal product, the process
works efficiently only if a specific range of external energy is provided.
82 |
Virginia Tech | Figure 3.13 Schematic showing behavior of released water droplets from the coal
agglomerates in liquid pentane column (a) State I (b) State II (c) State III (d) State IV
The calculated free energies are theoretical values, which cannot be easily verified, but
these literature values provide some insight into how the process works. It is now necessary to
provide scientific evidence, which supports the conclusions from proposed theory.
3.6.1 Investigation for Water Coalescence Mechanism
The coalescence mechanism of small water droplets released during the de-agglomeration
process into larger water droplets is understood thermodynamically. To support the theory, a
simple investigation was conducted using fluorescein, which is only soluble with water.
Agglomerates were prepared with fluorescein mixed into ultrafine coal slurry and screened. The
testing apparatus includes a glass beaker filled with liquid pentane and a dispersion device. The
agglomerates, in small batches, were then poured constantly into the pentane bath in the beaker
equipped with a mixer agitating at a low speed (80 RPM). To investigate, a video file was
recorded for the complete experiment and analyzed.
Figure 3.14 shows still images from the recorded video in an incremental order of time.
At 30 seconds, agglomerates were poured from the top and quickly started dispersing in pentane
as illustrated in Figure 3.14(a) and 3.14(b). After 2 minutes, signs of fluorescein dyed water,
83 |
Virginia Tech | analyzed. The photograph of the bottom of the beaker, shown in Figure 3.14 (g), clearly indicates
that the moisture droplets released from agglomerates eventually find themselves and coalesced
into a big water globule.
3.6.2 Investigation on Energy Input for De-agglomeration
In the current version of the HHS process, vibration energy is used as an external source
of energy. The newly developed vibrating mesh device creates a uniform hydrodynamic shear
field in the pentane column using vibrational energy. To determine if the moisture separation
process works efficiently in a specific range of energy, criteria for the vibration energy need to
be defined experimentally.
From the bench-scale experimental studies, two dimensionless operating parameters
related to vibration energy are considered to be critical for effective dispersion and dewatering of
coal particles. The first, which is called “vibration strength” (ζ), is defined as the ratio of the
vibration force to the gravitational force. Vibration strength, also called vibration number, is a
widely recognized parameter (Pakowaski et. al., 1984; Daleffe et al., 2004; Levy et al., 2006;
Meili et al., 2010) used to quantify vibration energy in studies of vibro-fluidized beds for drying
technologies. Mathematically, vibration strength can be expressed in terms of vibrational
frequency (f) and amplitude of vibrations (A) as illustrated in Equation 3.7:
[3.6]
( )
where ω is angular frequency of mechanical vibration. Since ω=2πf, Equation 3.6 can be re-
written as:
( ) ( )
[3.7]
85 |
Virginia Tech | Figure 3.15 Effect of Vibration strength on product moisture
The second operating parameter is the dimensionless length, which is the ratio of
vibrational amplitude (A) and the pentane column length (Z ) in the reactor. It was observed that
p
these parameters affect the moisture separation individually as well as in the combination with
each other. The detailed description of the interactions between these parameters is provided in
Chapter 4, as these criteria were critical in developing the reactor for the Proof-of-Concept
(POC) pilot-scale unit.
Figure 3.15 is a semi-log plot between product moisture and the vibrational strength. The
plot shows the variation in product moisture contents as a function of vibration strength for
different dimensionless lengths. When low strength is applied (ζ < 2.5), the process produced
consistently low moisture products (<4%). When a higher strength is applied (ζ > 10), the
moisture in the product varied over a wide range and was controlled primarily by other
parameters. The increase in moisture indicates that an excessive amount of energy may hinder
the water coalescence mechanism inside the reactor. It is very evident from the data that the
increased in energy may not necessarily improve the kinetics of dispersion; rather it can
86 |
Virginia Tech | substantially hamper the product quality. From engineering point of view, this information is
highly vital both in terms of economics and scale-up.
3.7 Conclusions
The HHS process implements the novel concept of dewatering by displacement. The
process makes use of a novel vibrating mesh reactor, which is the crux of the proposed
technology. The reactor substantially improves coal quality after oil-agglomeration of ultrafine
coal particles by dispersing coal particles in the oil phase and by coalescing and rejecting water
droplets that would otherwise be trapped in the coal-oil agglomerates. In order to better
understand the technology, several series of bench-scale tests were conducted to provide insight
regarding the underlying thermodynamics and kinetics that control the system. The
thermodynamic studies uncovered several important aspects of the process, including the water
coalescence mechanism and the specific range of energy required for an efficient performance.
Likewise, the kinetic studies assisted in determining the rates of homogenization and de-
agglomeration, which directly control the behavior of particles inside the reactor. The
information obtained from the thermodynamic and kinetic studies is crucial for the design of the
reactor.
References
1. Bethell, P.J. and Barbee, C.J. (2007), “Today’s Coal Preparation Plant: A Global
Perspective”, Designing the Coal Preparation Plant of the Future edited by Arnold, B.,
Klima, M. and Bethell, P., Pages 9-20, Society for Mining, Metallurgy, and Exploration.
2. Bika, D.G., Gentzler, M. and Michaels, J.N. (2001), “Mechanical Properties of
Agglomerates”, Powder Technology, Volume 117, Page 98.
3. Binks, B.P. (2002), Current Opinion, Colloid Interface Science, Volume -7, Page 21-41.
4. Boyle, J.F., Zloczower, I.M. and Feke, D.L. (2005), “Hydrodynamic Analysis of the
Mechanisms of Agglomerate Dispersion”, Powder Technology, Volume 153, Pages 127-
133.
5. Capes, C.E., McIIhinney, A.E., Russell, D.S. and Sirianni, A.F. (1974), “Rejection of
Trace Metals from Coal during Beneficiation by Agglomeration”, Environmental Science
and Technology, Volume 8, Pages 35-38.
87 |
Virginia Tech | CHAPTER 4 – Engineering Development of HHS Process POC Pilot Plant
4.1 Introduction
After successful bench-scale demonstration of the Hydrophobic-Hydrophilic Separation
(HHS) process, a preliminary design for a proof-of-concept (POC) pilot plant was developed
using the batch-scale testing data. Perigon, a chemical engineering design firm based in North
Carolina, was consulted for review of the preliminary designs and for recommendations on
process equipment and safety features for the POC plant. The construction of the POC pilot plant
was started at the Department of Mining and Minerals Research Laboratory in Virginia Tech and
was sponsored by Evan Energy, LLC, an investment company based in Richmond, Virginia. The
POC processing unit was designed for cleaning fine coal (0.15 mm x 0) with a raw dry feed
capacity of 100 pounds/hour. To take advantage of gravity flow and to minimize the pumping
requirement in the POC operation, a modular type design was developed.
The construction of the pilot plant started in August 2012 and completed in June 2013.
The POC pilot scale units were constructed as totally non-electric, enclosed units due to the
presence of flammable hydrocarbon liquid. Power required for agitating the slurry and moving
the material through the process was provided by pneumatic motors. For pumping the slurry,
peristaltic pumps were utilized, which eliminated any possible leaks due to pump shaft seals. All
the processing units were designed to be operated under a nitrogen blanket at a slightly elevated
pressure. All the necessary safety features were incorporated in the POC pilot plant. A multi-
point pentane and oxygen sensing system was installed to monitor the leaks at the agitator shaft
seals, in the vent from the condenser, and near the floor around the processing unit. Ancillary
electrical equipment supporting the processing unit (water heater, water chiller, nitrogen
generator, compressed air supply, etc.) was located in an adjacent permanent building, while the
POC plant was constructed in the open-air shed. To eliminate any possibility of static electricity
due to the fluid flow, all process units and streams were grounded with multiple ground rods. In
addition, a series of startup, shutdown and operating protocols were developed for a safe
operation.
Shakedown testing commenced in June 2013 and was completed in August 2013. Several
modifications were implemented in the POC operation during the period, which are discussed in
the chapter.
90 |
Virginia Tech | 4.2 Development of Process Flowsheet Diagram (PFD)
After the successful bench-scale studies a 100 lbs/hr process flowsheet for the HHS
process was developed using the data obtained from the laboratory testing. The LIMN® process
flowsheet software developed by David Wiseman Pty Ltd was used for balancing solids, water,
and pentane flow rates under steady-state conditions. LIMN® is a Microsoft Excel-based
software package, which has built in routines for process modeling.
In the proposed flowsheet, several parameters were examined including pentane losses
with the discharge of solid products and losses due to pentane solubility in water. Figure 4.1
shows the balanced flowsheet designed for a clean coal product with a target moisture of 6%. In
the proposed design, two new units were added in the process, which were not implemented
earlier in the continuous bench-scale set-up. These units were a sieve bend to minimize water
flow into the phase separator (vibrating mixer) and a clarifying thickener to minimize pentane
and carbon losses.
In the developed flowsheet, the feed stream is treated with a high dosage of liquid n-
pentane (25% by weight of carbon content) in a high/low shear mixing tank for agglomeration
process. The agglomerated overflow from the mixer unit passes through a sieve bend where the
sieve oversize material (mostly agglomerates) treated in a phase separator (later called as
“vibratory mixer”), which is the key unit operation in the process. The undersize stream from the
sieve bend, which mainly contains water and mineral matter, passes to a clarifying thickener.
The thickener underflow is rejected from the process as waste. Any residual pentane or partially
agglomerated fine coal particles that are lighter than water float as overflow from the thickener
and are recycled back to the high/low-shear mixer. This additional clarifying step helps in
minimizing the losses of pentane in the process.
The phase separator disperses the agglomerated coal and entrapped moisture (around
40% moisture) in pentane as described in the batch scale testing. The final target moisture was
specified 6% by weight in the flowsheet to balance the water in the process streams. The phase
separator overflow with 10% solids concentration (dry coal in pentane) is pumped to a
hydrophobic liquid (HL) thickener. Coal particles and associated moisture settle quickly in
pentane liquid because of the higher differential density between clean coal, water and pentane.
It was observed during the batch testing that some amount of moisture (4-8%) in product actually
facilitates coal settling in the HL thickener. The underflow of the phase separator, which mainly
91 |
Virginia Tech | contains water and mineral matter, is pumped into the clarifying thickener. The HL thickener is
initially filled with pentane prior to operation. The supernatant from HL thickener, which may
carry a small percentage of solids, is transferred back to the phase separator. This pentane
recycle helps in continuous flushing of dispersed coal from the phase separator to the HL
thickener. The settled coal from the HL thickener (30% solids) is pumped to a pentane recovery
circuit. The pentane recovery circuit is comprised of an evaporator and a condensing unit. The
condensed pentane liquid, which is pure pentane, is transferred back to the pure pentane drum. A
make-up pentane stream covers any losses of pentane in the process.
The LIMN® software was used to balance the solid and liquid flow rates through
the process circuit. The simulation required several thousand iterations to reach steady state. The
simulation results indicated that the total loss of pentane would be approximately 0.067 lbs/hr
based on pentane solubility in water and pentane absorption in coal (obtained from bench-scale
tests). This loss equates to about 1.34 pounds per ton of coal processed.
The simulated flowsheet proved to be very helpful for the engineering design and
development of a proof-of-concept (POC) pilot-scale plant, which is the focus of this research.
The flowsheet was further modified during the scale-up and shakedown testing as new data
became available, as will be discussed later in this document. Additionally, Perigon, a chemical
engineering firm based in North Carolina, was consulted regarding the appropriateness of the
simulated flowsheet. The consultation was primarily to review the POC design and to
recommend the pentane recovery system. Moreover, Perigon assisted in identifying safety
features necessary to operate the POC plant in an intrinsically safe manner.
4.3 Procedure for Development of POC Unit Operations
The primary intent to develop the POC pilot plant is to demonstrate the separation
capabilities of the HHS process on a large-scale. The HHS process is unique in that the
technology involves several sub-processes such as mixers, size separators, heat-exchangers, and
a novel dispersion/de-agglomeration method for dewatering. The engineering development of the
POC unit operations was not only limited to technical data obtained from bench-scale testing, but
also on scientific and engineering judgments based on previous research in the published
literature.
93 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.